HEALTH TECHNOLOGY ASSESSMENT

Transcription

1 HEALTH TECHNOLOGY ASSESSMENT VOLUME 22 ISSUE 37 JUNE 2018 ISSN A pedometer-based waking intervention in 45- to 75-year-ods, with and without practice nurse support: the PACE-UP three-arm custer RCT Tess Harris, Say Kerry, Christina Victor, Steve Iiffe, Michae Ussher, Juia Fox-Rushby, Peter Whincup, Uf Ekeund, Chery Furness, Eizabeth Limb, Nana Anokye, Judith Ibison, Stephen DeWide, Lee David, Emma Howard, Rebecca Dae, Jaime Smith, Rebecca Normanse, Caroe Beighton, Katy Morgan, Charotte Wahich, Sabina Sanghera and Derek Cook DOI /hta22370

2

3 A pedometer-based waking intervention in 45- to 75-year-ods, with and without practice nurse support: the PACE-UP three-arm custer RCT Tess Harris, 1 * Say Kerry, 2 Christina Victor, 3 Steve Iiffe, 4 Michae Ussher, 1 Juia Fox-Rushby, 5 Peter Whincup, 1 Uf Ekeund, 6,7 Chery Furness, 1 Eizabeth Limb, 1 Nana Anokye, 5 Judith Ibison, 1 Stephen DeWide, 1 Lee David, 8 Emma Howard, 1 Rebecca Dae, 1 Jaime Smith, 1 Rebecca Normanse, 1 Caroe Beighton, 1 Katy Morgan, 2 Charotte Wahich, 1 Sabina Sanghera 5 and Derek Cook 1 1 Popuation Heath Research Institute, St George s, University of London, London, UK 2 Pragmatic Cinica Trias Unit, Queen Mary University of London, London, UK 3 Gerontoogy and Heath Services Research Unit, Brune University London, London, UK 4 Research Department of Primary Care and Popuation Heath, University Coege London, London, UK 5 Heath Economics Research Group, Brune University London, London, UK 6 Department of Sports Medicine, Norwegian Schoo of Sport Sciences, Oso, Norway 7 Medica Research Counci Epidemioogy Unit, University of Cambridge, Cambridge, UK 8 10 Minute CBT, Devonshire Business Centre, Letchworth Garden City, UK *Corresponding author Decared competing interests of authors: Lee David is the director of 10 Minute CBT. She heped to deveop Pedometer And Consutation Evauation-UP (PACE-UP) patient resources and training for the PACE-UP nurses and received persona fees from 10 Minute CBT during the conduct of this study. Tess Harris is a member of the Nationa Institute for Heath Research (NIHR) Heath Technoogy Assessment Primary Care and Community Preventative Interventions pane. Juia Fox-Rushby reports grants from the NIHR and membership of the Pubic Heath Research Research Funding Board during the conduct of the study. Katy Morgan s saary was funded through a NIHR research methods feowship (reference number MET-12-16), during the conduct of the study.

4

5 Pubished June 2018 DOI: /hta22370 This report shoud be referenced as foows: Harris T, Kerry S, Victor C, Iiffe S, Ussher M, Fox-Rushby J, et a. A pedometer-based waking intervention in 45- to 75-year-ods, with and without practice nurse support: the PACE-UP three-arm custer RCT. Heath Techno Assess 2018;22(37). Heath Technoogy Assessment is indexed and abstracted in Index Medicus/MEDLINE, Excerpta Medica/EMBASE, Science Citation Index Expanded (SciSearch ) and Current Contents / Cinica Medicine.

6

7 Heath Technoogy Assessment HTA/HTA TAR ISSN (Print) ISSN (Onine) Impact factor: Heath Technoogy Assessment is indexed in MEDLINE, CINAHL, EMBASE, The Cochrane Library and the Carivate Anaytics Science Citation Index. This journa is a member of and subscribes to the principes of the Committee on Pubication Ethics (COPE) ( Editoria contact: journas.ibrary@nihr.ac.uk The fu HTA archive is freey avaiabe to view onine at Print-on-demand copies can be purchased from the report pages of the NIHR Journas Library website: Criteria for incusion in the Heath Technoogy Assessment journa Reports are pubished in Heath Technoogy Assessment (HTA) if (1) they have resuted from work for the HTA programme, and (2) they are of a sufficienty high scientific quaity as assessed by the reviewers and editors. Reviews in Heath Technoogy Assessment are termed systematic when the account of the search appraisa and synthesis methods (to minimise biases and random errors) woud, in theory, permit the repication of the review by others. HTA programme The HTA programme, part of the Nationa Institute for Heath Research (NIHR), was set up in It produces high-quaity research information on the effectiveness, costs and broader impact of heath technoogies for those who use, manage and provide care in the NHS. Heath technoogies are broady defined as a interventions used to promote heath, prevent and treat disease, and improve rehabiitation and ong-term care. The journa is indexed in NHS Evidence via its abstracts incuded in MEDLINE and its Technoogy Assessment Reports inform Nationa Institute for Heath and Care Exceence (NICE) guidance. HTA research is aso an important source of evidence for Nationa Screening Committee (NSC) poicy decisions. For more information about the HTA programme pease visit the website: This report The research reported in this issue of the journa was funded by the HTA programme as project number 10/32/02. The contractua start date was in March The draft report began editoria review in June 2017 and was accepted for pubication in November The authors have been whoy responsibe for a data coection, anaysis and interpretation, and for writing up their work. The HTA editors and pubisher have tried to ensure the accuracy of the authors report and woud ike to thank the reviewers for their constructive comments on the draft document. However, they do not accept iabiity for damages or osses arising from materia pubished in this report. This report presents independent research funded by the Nationa Institute for Heath Research (NIHR). The views and opinions expressed by authors in this pubication are those of the authors and do not necessariy refect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Heath and Socia Care. If there are verbatim quotations incuded in this pubication the views and opinions expressed by the interviewees are those of the interviewees and do not necessariy refect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Heath and Socia Care. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Pubished by the NIHR Journas Library ( produced by Prepress Projects Ltd, Perth, Scotand (

8 NIHR Journas Library Editor-in-Chief Professor Tom Waey Director, NIHR Evauation, Trias and Studies and Director of the EME Programme, UK NIHR Journas Library Editors Professor Ken Stein Chair of HTA and EME Editoria Board and Professor of Pubic Heath, University of Exeter Medica Schoo, UK Professor Andrée Le May Chair of NIHR Journas Library Editoria Group (HS&DR, PGfAR, PHR journas) Dr Martin Ashton-Key Consutant in Pubic Heath Medicine/Consutant Advisor, NETSCC, UK Professor Matthias Beck Professor of Management, Cork University Business Schoo, Department of Management and Marketing, University Coege Cork, Ireand Dr Tessa Criy Director, Crysta Bue Consuting Ltd, UK Dr Eugenia Cronin Senior Scientific Advisor, Wessex Institute, UK Dr Peter Davidson Director of the NIHR Dissemination Centre, University of Southampton, UK Ms Tara Lamont Scientific Advisor, NETSCC, UK Dr Catriona McDaid Senior Research Feow, York Trias Unit, Department of Heath Sciences, University of York, UK Professor Wiiam McGuire Professor of Chid Heath, Hu York Medica Schoo, University of York, UK Professor Geoffrey Meads Professor of Webeing Research, University of Winchester, UK Professor John Norrie Chair in Medica Statistics, University of Edinburgh, UK Professor John Powe Consutant Cinica Adviser, Nationa Institute for Heath and Care Exceence (NICE), UK Professor James Raftery Professor of Heath Technoogy Assessment, Wessex Institute, Facuty of Medicine, University of Southampton, UK Dr Rob Riemsma Reviews Manager, Keijnen Systematic Reviews Ltd, UK Professor Heen Roberts Professor of Chid Heath Research, UCL Great Ormond Street Institute of Chid Heath, UK Professor Jonathan Ross Professor of Sexua Heath and HIV, University Hospita Birmingham, UK Professor Heen Snooks Professor of Heath Services Research, Institute of Life Science, Coege of Medicine, Swansea University, UK Professor Jim Thornton Professor of Obstetrics and Gynaecoogy, Facuty of Medicine and Heath Sciences, University of Nottingham, UK Professor Martin Underwood Director, Warwick Cinica Trias Unit, Warwick Medica Schoo, University of Warwick, UK Pease visit the website for a ist of editors: Editoria contact: journas.ibrary@nihr.ac.uk NIHR Journas Library

9 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Abstract A pedometer-based waking intervention in 45- to 75-year-ods, with and without practice nurse support: the PACE-UP three-arm custer RCT Tess Harris, 1 * Say Kerry, 2 Christina Victor, 3 Steve Iiffe, 4 Michae Ussher, 1 Juia Fox-Rushby, 5 Peter Whincup, 1 Uf Ekeund, 6,7 Chery Furness, 1 Eizabeth Limb, 1 Nana Anokye, 5 Judith Ibison, 1 Stephen DeWide, 1 Lee David, 8 Emma Howard, 1 Rebecca Dae, 1 Jaime Smith, 1 Rebecca Normanse, 1 Caroe Beighton, 1 Katy Morgan, 2 Charotte Wahich, 1 Sabina Sanghera 5 and Derek Cook 1 1 Popuation Heath Research Institute, St George s, University of London, London, UK 2 Pragmatic Cinica Trias Unit, Queen Mary University of London, London, UK 3 Gerontoogy and Heath Services Research Unit, Brune University London, London, UK 4 Research Department of Primary Care and Popuation Heath, University Coege London, London, UK 5 Heath Economics Research Group, Brune University London, London, UK 6 Department of Sports Medicine, Norwegian Schoo of Sport Sciences, Oso, Norway 7 Medica Research Counci Epidemioogy Unit, University of Cambridge, Cambridge, UK 8 10 Minute CBT, Devonshire Business Centre, Letchworth Garden City, UK *Corresponding author tharris@sgu.ac.uk Background: Guideines recommend waking to increase moderate to vigorous physica activity (MVPA) for heath benefits. Objectives: To assess the effectiveness, cost-effectiveness and acceptabiity of a pedometer-based waking intervention in inactive aduts, deivered postay or through dedicated practice nurse physica activity (PA) consutations. Design: Parae three-arm tria, custer randomised by househod. Setting: Seven London-based genera practices. Participants: A tota of 11,015 peope without PA contraindications, aged years, randomy seected from practices, were invited. A tota of 6399 peope were non-responders, and 548 peope sef-reporting achieving PA guideines were excuded. A tota of 1023 peope from 922 househods were randomised to usua care (n = 338), posta intervention (n = 339) or nurse support (n = 346). The recruitment rate was 10% (1023/10,467). A tota of 956 participants (93%) provided outcome data. Interventions: Intervention groups received pedometers, 12-week waking programmes advising participants to graduay add 3000 steps in 30 minutes most days weeky and PA diaries. The nurse group was offered three dedicated PA consutations. Main outcome measures: The primary and main secondary outcomes were changes from baseine to 12 months in average daiy step counts and time in MVPA (in 10-minute bouts), respectivey, from 7-day acceerometry. Individua resource-use data informed the within-tria economic evauation and the Markov Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. vii

10 ABSTRACT mode for simuating ong-term cost-effectiveness. Quaitative evauations assessed nurse and participant views. A 3-year foow-up was conducted. Resuts: Baseine average daiy step count was 7479 [standard deviation (SD) 2671], average minutes per week in MVPA bouts was 94 minutes (SD 102 minutes) for those randomised. PA increased significanty at 12 months in both intervention groups compared with the contro group, with no difference between interventions; additiona steps per day were 642 steps [95% confidence interva (CI) 329 to 955 steps] for the posta group and 677 steps (95% CI 365 to 989 steps) for nurse support, and additiona MVPA in bouts (minutes per week) was 33 minutes per week (95% CI 17 to 49 minutes per week) for the posta group and 35 minutes per week (95% CI 19 to 51 minutes per week) for nurse support. Intervention groups showed no increase in adverse events. Incrementa cost per step was 19p and 3.61 per minute in a 10-minute MVPA bout for nurse support, whereas the posta group took more steps and cost ess than the contro group. The posta group had a 50% chance of being cost-effective at a 20,000 per quaity-adjusted ife-year (QALY) threshod within 1 year and had both ower costs [ 11M (95% CI 12M to 10M) per 100,000 popuation] and more QALYs [759 QALYs gained (95% CI 400 to 1247 QALYs)] than the nurse support and contro groups in the ong term. Participants and nurses found the interventions acceptabe and enjoyabe. Three-year foow-up data showed persistent intervention effects (nurse support pus posta vs. contro) on steps per day [648 steps (95% CI 272 to 1024 steps)] and MVPA bouts [26 minutes per week (95% CI 8 to 44 minutes per week)]. Limitations: The 10% recruitment eve, with ower eves in Asian and socioeconomicay deprived participants, imits the generaisabiity of the findings. Assessors were unmasked to the group. Concusions: A primary care pedometer-based waking intervention in 45- to 75-year-ods increased 12-month step counts by around one-tenth, and time in MVPA bouts by around one-third, with simiar effects for the nurse support and posta groups, and persistent 3-year effects. The posta intervention provides cost-effective, ong-term quaity-of-ife benefits. A primary care pedometer intervention deivered by post coud hep address the pubic heath physica inactivity chaenge. Future work: Exporing different recruitment strategies to increase uptake. Integrating the Pedometer And Consutation Evauation-UP (PACE-UP) tria with evoving PA monitoring technoogies. Tria registration: Current Controed Trias ISRCTN Funding: This project was funded by the Nationa Institute for Heath Research (NIHR) Heath Technoogy Assessment programme and wi be pubished in fu in Heath Technoogy Assessment; Vo. 22, No. 37. See the NIHR Journas Library website for further project information. viii NIHR Journas Library

11 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Contents List of tabes List of figures List of abbreviations Pain Engish summary Scientific summary xv xix xxi xxiii xxv Chapter 1 Introduction: why this study was needed 1 Benefits and risks of physica activity and current physica activity guideines 1 What are the benefits of physica activity for aduts and oder aduts? 1 What are the current physica activity guideines and who is achieving them? 1 What are the risks from increasing physica activity? 1 Strategies for increasing physica activity 2 How can aduts and oder aduts increase their physica activity eves? 2 Are pedometers hepfu? 2 How do step count goas reate to physica activity recommendations? 3 Coud acceerometers be usefu in a pedometer-based waking intervention? 3 Are pedometers cost-effective? 3 What is the roe of primary care in promoting physica activity? 3 Theoretica base, pioting and preparatory work to deveop the intervention 4 Rationae for research 4 Chapter 2 Methods 7 Study design 7 Study aims and objectives 7 Study aims 7 Primary objectives (reating to the primary outcome of step counts) 7 Secondary objectives (reating to secondary outcomes of time in moderate to vigorous physica activity in bouts, sedentary time and cost-effectiveness) 8 Other objectives 8 Practice and participant incusion/excusion criteria 8 Practice incusion criteria 8 Participant incusion criteria 9 Participant excusion criteria 9 Recruitment of practices and participants: informed consent 9 Practice recruitment 9 Participant recruitment 9 Non-responders and non-participants 10 Informed consent 10 Changes from the pubished protoco 10 Interventions 10 Procedure for the posta intervention group 11 Procedure for the nurse intervention group 11 Procedure for the contro group 12 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. ix

12 CONTENTS Outcome measures 13 Primary and secondary outcome measures 13 Anciary outcomes 15 Ascertainment of outcomes 15 Acceerometry 15 Anthropometry 15 Questionnaire measures 15 Primary care computerised record measures 16 Baseine and foow-up data coection 16 Baseine data coection 16 Foow-up data coection 16 Acceerometer data reduction 17 Adverse events and serious adverse events 17 Sampe size 18 Randomisation, conceament of aocation, contamination and treatment masking 18 Randomisation and conceament of aocation 18 Contamination 19 Treatment masking 19 Withdrawas, osses to foow-up and missing data 19 Withdrawas and osses to foow-up 19 Procedure for accounting for missing data 19 Statistica methods 20 Primary anaysis 20 Secondary and anciary outcome anayses 20 Adverse event anayses 21 Subgroup anayses 21 Sensitivity anayses 21 Ethics approva and research governance 21 Management of the tria 21 Further tria foow-up at 3 years 22 Chapter 3 Resuts 23 Recruitment of participants 23 Baseine characteristics of the study popuation (Tabe 4) 23 Losses to foow-up 23 Data competeness for acceerometry 27 Effect of the intervention on acceerometer-assessed physica activity outcomes (Tabe 5) 27 Three-month (interim) outcomes 27 Tweve-month (main) outcomes 28 Effect of the intervention on sef-reported physica activity outcomes at 12 months (Tabe 6) 28 Effect of the intervention on other heath-reated outcomes (Tabe 7) 29 Effect of the intervention on adverse events and serious adverse events (Tabe 8) 29 Subgroup anayses (Figure 4) 29 Sensitivity anayses and imputations (see Appendix 2, Tabe 26) 29 Summary of the main tria findings 32 Chapter 4 Economic evauation 35 Introduction 35 Within-tria cost-effectiveness anaysis 35 Methods 36 Resuts 38 x NIHR Journas Library

13 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Beyond-tria modeing 42 Methods 43 Resuts 44 Discussion 46 Strengths and weaknesses 47 Concusion 47 Chapter 5 Generaisabiity 49 Introduction 49 Methods 49 Data coection for quantitative comparisons 49 Comparison groups 49 Statistica anaysis 50 Methods for the interview study of non-participants 51 Methods for anaysing interviews 51 Resuts 51 Resuts from quantitative comparisons 51 Resuts from quaitative comparisons 56 Summary of the findings 59 Strengths and imitations 59 Comparisons with previous work 60 Impications 61 Concusions 62 Chapter 6 Process evauation 63 Introduction 63 Why is process evauation necessary in the PACE-UP tria? 63 Medica Research Counci s guidance on process evauations in compex interventions 63 Methods and resuts 66 Impementation 67 Impementation process 67 Reach 69 Fideity (content and quaity) 69 Dose 72 Adaptations 73 Mechanisms of impact 73 Participant responsiveness 73 Context 74 Association between process evauation measures and tria outcome measures 75 Nurse session attendance and physica activity outcomes 75 Diary return and physica activity outcomes 75 Pedometer use and physica activity outcomes 75 Discussion 76 Main findings 76 Strengths and imitations 77 Comparison with other compex intervention process evauations 78 Impications of the process evauation for the interpretation of the PACE-UP tria 78 Concusion 79 Chapter 7 What did the nurses and participants think about the intervention? 81 Introduction 81 Recruiting tria participants for the quaitative study 81 The roe of nurses in the PACE-UP intervention 82 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xi

14 CONTENTS Preparing for the tria 83 Deivering and participating in the tria 83 Perceived vaue of nurse consutations 83 Behaviour change techniques 84 Adapting the tria protoco versus fideity 85 Tria materias and equipment 86 Nurse satisfaction 86 After the tria and impementation 87 Discussion 88 Chapter 8 Three-year foow-up to assess the maintenance of physica activity eves 89 Introduction 89 Methods for quantitative physica activity evauation 90 Ethics approva and research governance 90 Participants eigibe for the 3-year foow-up 90 Contacting participants 90 Informed consent 90 Data coection 91 Outcome measures 91 Acceerometer data reduction 91 Procedure for accounting for missing data 91 Statistica methods 91 Methods for quaitative evauation 92 Samping 92 Recruitment and informed consent 92 Transcribing 92 Interview schedue 93 Anaysis 93 Resuts for the quantitative physica activity evauation 93 Foow-up rate 93 Data competeness 93 Objective physica activity findings 93 Missing data anayses 93 Resuts for the quaitative evauation 96 Factors affecting physica activity eves and maintenance at 3 years 97 The effect of the minima intervention on physica activity eves of participants in the contro group 100 Discussion 101 Main quantitative findings from the 3-year foow-up 101 Main quaitative findings from the 3-year foow-up interviews 102 Chapter 9 Discussion 103 Summary of the findings 103 Strengths and imitations 104 Study strengths 104 Study imitations 104 Generaisabiity 105 Comparison with other studies 105 Interpretation of the resuts 107 Concusions 108 Impications for heath care 108 Recommendations for research 108 xii NIHR Journas Library

15 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Acknowedgements 109 References 113 Appendix 1 Methods 129 Appendix 2 Resuts 207 Appendix 3 Economic evauation 213 Appendix 4 Generaisabiity 233 Appendix 5 Process evauation 243 Appendix 6 Quaitative evauation 257 Appendix 7 Three-year foow-up 261 Appendix 8 Discussion 273 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xiii

16

17 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 List of tabes TABLE 1 Components of the compex intervention for the PACE-UP tria 11 TABLE 2 The PACE-UP patient handbook, and diary and BCTs incuded 12 TABLE 3 The PACE-UP practice nurse PA consutations and BCTs incuded 13 TABLE 4 Baseine characteristics of the 1023 randomised subjects 25 TABLE 5 Primary and secondary acceerometry outcome data 27 TABLE 6 Effect estimates for sef-report questionnaires (IPAQ and GPPAQ) at 12 months 28 TABLE 7 Effect estimates for other heath-reated outcomes 30 TABLE 8 Adverse events 31 TABLE 9 Average costs and QALYs per participant, by tria arm ( , base case, with missing vaues imputed) 39 TABLE 10 Costs, effects and cost-effectiveness at 3 and 12 months ( ; base case, adjusted for baseine differences) 39 TABLE 11 Costs, effects and cost-effectiveness over a ifetime (cohort of 100,000) 45 TABLE 12 Responders to the invitation etter by age, sex and IMD quintie 52 TABLE 13 Competion of the baseine assessment and questionnaires in participants and non-participants who are not active on the GPPAQ, by age, sex, IMD quintie and ethnicity 53 TABLE 14 Participants and non-participants who competed questionnaires and were not active on the GPPAQ: demographics and heath and ifestye factors 54 TABLE 15 Summary of categories and themes: barriers to participating 56 TABLE 16 Summary of the PACE-UP tria process evauation data sources, evauative groups and reported measures 66 TABLE 17 Quaity of deivery and participant responsiveness data from the nurse and patient aiance questionnaires 70 TABLE 18 Fideity: quaity scores of performance for audio-recordings of nurse sessions by BCT trainer 72 TABLE 19 The PACE-UP tria modeing resuts: reating nurse session attendance, step count diary return and pedometer use to PA outcomes 76 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xv

18 LIST OF TABLES TABLE 20 Summary means and SDs for acceerometry data at baseine, 3 months, 12 months and 3 years 95 TABLE 21 Acceerometry outcome data at 3 months, 12 months and 3 years. Anaysed using a avaiabe data at each foow-up. N = 954 at 3 months, N = 956 at 12 months and N = 681 at 3 years 96 TABLE 22 Imputation anayses for the 3-year acceerometry outcomes 97 TABLE 23 The number of days with 540 minutes acceerometer wear time by treatment group at baseine, 3 months and 12 months 209 TABLE 24 Summary data for main outcome and anciary outcome variabes 210 TABLE 25 Physica activity measured by sef-report IPAQ and GPPAQ measurements 211 TABLE 26 Sensitivity and imputation anayses for the primary outcome (step count at 12 months) 212 TABLE 27 Resource use and cost components of set-up costs 213 TABLE 28 Components of the cost of deivering care and sources of data by tria arm 215 TABLE 29 Heath provider cost of heath service use 215 TABLE 30 Persona costs of participating in the interventions or contro arms of the tria 216 TABLE 31 Resource use and cost components of set-up cost 216 TABLE 32 Components of deivery cost of intervention (posta deivery group) 219 TABLE 33 Components of deivery cost of intervention (nurse-support group) 219 TABLE 34 Costs to participants of participating in interventions and PA 220 TABLE 35 Heath service use by tria arm with unit costs 221 TABLE 36 Within-tria sensitivity anayses (at 12 months) 226 TABLE 37 Parameter vaues for the ong-term cost-effectiveness mode 229 TABLE 38 Impementation process: training deivered to nurses 243 TABLE 39 Fideity: content deivered in nurse intervention group sessions in accordance with nurse checkists 243 TABLE 40 Nurse-support and posta intervention group PA diary return and use 244 xvi NIHR Journas Library

19 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 41 Pedometer use from questionnaire data 244 TABLE 42 Duration of sessions 244 TABLE 43 Interview participant detais 259 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xvii

20

21 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 List of figures FIGURE 1 A summary of the PACE-UP waking programme 14 FIGURE 2 Schedue of outcome assessment measures used in the PACE-UP tria 17 FIGURE 3 The PACE-UP tria CONSORT fow diagram 24 FIGURE 4 Subgroup anayses 32 FIGURE 5 Cost-effectiveness pane for the posta deivery group vs. the contro group at 12 months 41 FIGURE 6 Cost-effectiveness pane for the nurse-support group vs. the contro group at 12 months 42 FIGURE 7 Cost-effectiveness acceptabiity curve for the posta deivery and nurse-support groups (vs. the contro group) at different wiingness-to-pay-per- QALY threshods 42 FIGURE 8 Iustration of pathways within the ong-term cost-effectiveness mode 43 FIGURE 9 Cost-effectiveness acceptabiity curve showing the probabiity of ifetime cost-effectiveness for the posta deivery group and the nurse-support group (vs. the contro group) at different wiingness-to-pay threshod eves 45 FIGURE 10 Fow chart to show the recruitment process in the PACE-UP tria 50 FIGURE 11 The key functions of process evauation and the reationships between these 64 FIGURE 12 Logic mode for the PACE-UP (Pedometers and Consutation Evauation UP) PA tria 65 FIGURE 13 The key functions of process evauation and the reationships between them for the PACE-UP tria 68 FIGURE 14 Behaviour change technique competency eve 71 FIGURE 15 The PACE-UP CONSORT fow diagram with 3-year foow-up data 94 FIGURE 16 Sensitivity anayses for different vaues of missing steps counts 98 FIGURE 17 Tria procedures and compex intervention components 207 FIGURE 18 Residuas from the 12-month modes for steps and weeky MVPA in 10-minute bouts 208 FIGURE 19 Cost-effectiveness pane for nurse support vs. posta deivery at 12 months 225 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xix

22 LIST OF FIGURES FIGURE 20 Cost-effectiveness acceptabiity curve showing the probabiity of within-tria cost-effectiveness for nurse support vs. posta deivery at different wiingness-to-pay threshod eves 225 FIGURE 21 Cost-effectiveness acceptabiity curve showing the probabiity of ifetime cost-effectiveness for the nurse support group compared with the posta deivery group at different wiingness-to-pay threshod eves 231 xx NIHR Journas Library

23 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 List of abbreviations A&E accident and emergency MRC Medica Research Counci AE BCT BMI CEAC CHD CI adverse event behaviour change technique body mass index cost-effectiveness acceptabiity curve coronary heart disease confidence interva MVPA NICE NIHR NS-SEC moderate to vigorous physica activity Nationa Institute for Heath and Care Exceence Nationa Institute for Heath Research Nationa Statistics Socio-economic Cassification CONSORT CVD EQ-5D Consoidated Standards of Reporting Trias cardiovascuar disease EuroQo-5 Dimensions OR PA PACE-LIFT odds ratio physica activity Pedometer Acceerometer Evauation-LIFT EQ-5D-3L EuroQo-5 Dimensions, three-eve version PACE-UP Pedometer And Consutation Evauation-UP EQ-5D-5L GP GPPAQ HADS EuroQo-5 Dimensions, five-eve version genera practitioner Genera Practice Physica Activity Questionnaire Hospita Anxiety and Depression Scae PAI PSA QALY RCT RR SAE physica activity index probabiistic sensitivity anaysis quaity-adjusted ife-year randomised controed tria reative risk serious adverse event ICER incrementa cost-effectiveness ratio SD standard deviation IMD Index of Mutipe Deprivation TMG tria management group IPAQ Internationa Physica Activity Questionnaire TSC Tria Steering Committee Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxi

24

25 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Pain Engish summary Physica inactivity is common and causes i heath. Waking brisky enough to make you warm and increase breathing and heart rate, but aow conversation, is moderate-intensity physica activity. Brisk waking for 30 minutes most days is a good way to improve heath. Pedometers measure step counts and can increase physica activity eves, but few studies invoving pedometers have objectivey measured participants physica activity or incuded ong-term foow-up. The Pedometer And Consutation Evauation-UP (PACE-UP) tria recruited 1023 inactive 45- to 75-year-ods from seven South London practices, and randomised them to a usua physica activity (contro) group or to one of two intervention groups. The posta group participants were sent a pedometer, diary and 12-week pedometer-based waking programme, advising them to graduay add in 3000 steps or a 30-minute wak on 5 or more days weeky. The nurse-support group received the same materias through practice nurse physica activity consutations. Physica activity and participant-reported 12-month outcomes were compared with the beginning of the tria, aong with the costs of each tria group. A further 3-year foow-up was conducted and ong-term vaue for money was estimated. Both intervention groups significanty increased their waking (step counts and time in moderate-intensity physica activity) compared with contros, with no difference between nurse and posta groups. Interventions were safe and acceptabe to participants and nurses. There was no effect on body size, pain or depression, but the nurse-support group participants increased their confidence in their abiity to exercise. The 3-year foow-up found persistent positive effects of both interventions on physica activity eves. The posta intervention provided more vaue for money than the nurse-support group or the contro group in the short and ong term. A primary care pedometer intervention, deivered by post or with nurse support, coud provide an effective way to increase physica activity eves in aduts and oder aduts, with the posta route offering the most vaue for money. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxiii

26

27 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Scientific summary Background Physica activity (PA) heps aduts and oder aduts to remain heathy, and improves physica function and emotiona we-being. Inactivity is an important risk factor for mortaity and eads to high heath service costs. One way to achieve current nationa and internationa PA guideines for heath is by doing at east 30 minutes of moderate to vigorous physica activity (MVPA) in at east 10-minute bouts on 5 days weeky. However, a graded dose response reationship exists for PA and heath; therefore, for inactive peope, any PA increase is vauabe, as is decreasing sedentary time. Waking is the most common adut PA, and moderate-intensity waking approximates 100 steps per minute, so using a pedometer to add 3000 steps in 30 minutes onto habitua activity heps to achieve PA guideines. Systematic reviews of pedometer-based waking interventions show significant step count increases. However, studies were mainy sma, recruited vounteers and had short-term foow-up. In addition, previous pedometer studies have not rigorousy evauated their effectiveness with or without face-to-face support, and have used step counts, not MVPA, as the outcome. Programmes using personaised PA goas and behavioura strategies can achieve PA increases. Cochrane reviews have caed for PA interventions to incude objective PA measurement, adverse events reporting, comparisons of face-to-face interventions with remote interventions, onger foow-up and cost-effectiveness evauations. Primary care provides an idea context for PA interventions, aowing popuation-based samping, practice nurse invovement and continuity of care. Brief PA advice in primary care is advocated; however, more primary care PA trias are required. Objectives The research questions were: 1. Does a 3-month posta pedometer-based waking intervention increase PA (step count and time in MVPA in bouts) in inactive 45- to 75-year od primary care patients at 12-month foow-up? 2. Do dedicated practice nurse PA consutations provide additiona benefit? We aso present cost-effectiveness anayses and effects on patient-reported outcomes, anthropometric measures and adverse events. A quaitative evauation expored participant and practice nurse views. Longer-term foow-up was conducted at 3 years. Methods Design A three-arm parae-group, custer randomised tria, comparing a 3-month pedometer-based waking intervention, by post or with nurse-support, with usua care. Randomisation was by househod, aowing individuas and coupes to participate, in a 1 : 1 : 1 ratio. Participants and setting Recruitment was from seven ethnicay and sociay diverse, south London-based genera practice popuations, between September 2012 and October The 12-month foow-up was competed in October Eigibe patients were aged years, without contraindications to increasing MVPA. Excusions incuded care home residents and those with unsuitabe medica conditions. Random sampes of 400 eigibe househods per practice were seected; this process was repeated unti enough individuas were recruited. Individua invitations were posted. Those participants who reported achieving 150 minutes of MVPA Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxv

28 SCIENTIFIC SUMMARY weeky on a vaidated sef-report PA questionnaire were excuded. Anonymised demographic data were avaiabe through genera practice records for a those invited, enabing investigation of tria recruitment inequaities. Non-participants were invited to compete a questionnaire. Procedures and intervention Individua informed consent was obtained and baseine assessment undertaken prior to randomisation. Identica outcome assessments were conducted for a three groups. An acceerometer (GT3X+; ActiGraph LLC, Pensacoa, FL, USA) was used for baseine, and 3- and 12-month masked PA assessment of step counts and time in different PA intensities. The interventions incorporated behaviour change techniques (BCTs) and incuded individuaised step count and PA goas and the 3000 in 30 PA intensity message. The key intervention components were pedometers (SW-200 Yamax Digi-Waker; Yamasa Tokei Keiki Co. Ltd, Tokyo, Japan) to record individua step counts, a patient handbook, a PA diary (incuding an individua 12-week waking pan) and three individuay taiored practice nurse PA (10- to 20-minute) consutations (nurse-support group ony). The patient handbook and diary expained that adding 3000 steps per day (or a 30-minute wak) on 5 days weeky to their baseine, progressing over 12 weeks, woud hep to achieve PA guideines. BCTs, incuding goas and panning, sef-monitoring, feedback and encouraging socia support, were incuded in the handbook, diary and nurse consutations. Contro group participants were offered a pedometer, a handbook and a diary after the 12-month foow-up. Outcomes A primary and secondary PA outcomes were assessed by 7-day acceerometry measurements. The primary outcome was change in average daiy step count between baseine and 12 months. The secondary PA outcomes were changes in step counts between baseine and 3 months; changes in time spent weeky in MVPA in 10-minute bouts; and time spent being sedentary between baseine and 3 and 12 months. The other secondary outcome was cost-effectiveness. Anciary outcomes were: changes in anthropometry (body mass index, waist circumference, body fat) at 12 months changes in patient-reported outcomes exercise sef-efficacy, anxiety, depression [as measured via the Hospita Anxiety and Depression Scae (HADS)], heath-reated quaity of ife [as measured via the EuroQo-5 Dimensions, five-eve version (EQ-5D-5L)], pain and sef-reported PA variabes [as measured via the Internationa Physica Activity Questionnaire (IPAQ), short form and the Genera Practice Physica Activity Questionnaire (GPPAQ)] at 3 and 12 months adverse outcomes fas, injuries, fractures, cardiovascuar events and deaths assessed from tria monitoring procedures, 3- and 12-month questionnaires and primary care records. Sampe size A sampe of 993 (331 per group) was needed to detect the 1000 steps per day difference at 12 months, comparing any two groups, with 90% power, at a p-vaue of 0.01, aowing for househod custering and 15% attrition. Statistica anayses Acceerometry regression anayses were in two stages. Stage 1 estimated the average daiy step count at 12 months and at baseine, derived by using the same two-eve mode (eve 1, day within individua; eve 2, individua) in which daiy step counts were regressed on day order of wearing the acceerometer (from day 1 to day 7) and day of week, as fixed effects. At stage 2, the estimated 12-month average daiy step count was regressed on the estimated baseine average daiy step count, month of baseine acceerometry, age, sex, genera practice and treatment group, effectivey measuring the change in step count over 12 months. In this anaysis, eve 1 was individua and eve 2 was househod. MVPA in 10-minute bouts, sedentary time, wear time and 3-month outcomes were anaysed using identica approaches. The change in anthropometric and patient-reported outcomes was estimated using stage 2 modes. xxvi NIHR Journas Library

29 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Economic evauation Cost-effectiveness was estimated, from the NHS viewpoint, to generate the incrementa cost per change in step count, minutes of MVPA in 10-minute bouts and quaity-adjusted ife-years (QALYs). The probabiity of the interventions being cost-effective given different wiingness-to-pay vaues for QALYs and incrementa net benefit (difference between monetised benefit and costs of the intervention vs. the comparator) was cacuated. A Markov mode used the resuts to simuate ifetime cost-effectiveness. Deterministic and probabiistic sensitivity anayses were undertaken for short- and ong-term anayses. Process evauation Data were coected contemporaneousy with tria data coection, and associations between process measures and tria outcome measures were sought. Quaitative evauation Teephone interviews were conducted with nurse and posta participants, targeting some participants who had increased their PA and some who had not, to investigate their views of the intervention and the barriers to, and faciitators of, increasing PA eves. A practice nurse focus group session was conducted to understand practice nurses experience of deivering the intervention. Three-year foow-up Participant foow-up was conducted 3 years from baseine, incuding posta 7-day acceerometry, questionnaire and quaitative teephone interviews. The atter were carried out with randomy seected nurse and posta participants, to examine the factors affecting PA eves and maintenance of any increase in PA; and with contro participants, to see the effect of the 12-month minima intervention on PA eves. Resuts Of the 11,015 peope invited, 6399 did not respond, 548 sef-reported PA guideine achievements and were excuded and 10% (1023/10,467) were randomised. Participation rates were ower in men, younger subjects, those iving in deprived postcode areas and Asian patients. Back peope were equay ikey to participate as white peope. Baseine findings for a those randomised were as foows: average steps per day, 7479 steps [standard deviation (SD) 2671 steps]; and average minutes per week in MVPA of 10-minute bouts, 94 minutes (SD 102 minutes). Overa, 21% of participants (218/1023) achieved the PA guideines of 150 minutes of MVPA in bouts. A tota of 93% of participants (956/1023) were incuded in the 12-month primary anayses. At the interim 3-month outcome, both intervention groups had increased their steps per day from baseine compared with the contro group. Additiona steps per day were 692 steps [95% confidence interva (CI) 363 to 1020 steps; p < 0.001] for the posta group, and 1172 steps (95% CI 844 to 1501 steps; p < 0.001) for the nurse-support group. The difference between intervention groups was statisticay significant: 481 steps (95% CI 153 to 809 steps; p = 0.004). MVPA findings showed a simiar pattern: additiona MVPA in bouts (minutes per week) was 43 minutes (95% CI 26 to 60 minutes; p < 0.001) for the posta group, and 61 minutes (95% CI 44 to 78 minutes; p < 0.001) for the nurse-support group; the difference between intervention groups was 18 minutes (95% CI 1 to 35 minutes; p = 0.04). Sedentary time and acceerometer wear time were simiar between groups. For the primary outcome, both intervention groups increased their step counts from baseine to 12 months compared with contro participants; additiona steps per day were 642 steps (95% CI 329 to 955 steps; p < 0.001) for the posta group, and 677 steps (95% CI 365 to 989 steps; p < 0.001) for the nurse-support group, with no statisticay significant difference between intervention groups (36 steps, 95% CI 277 to 349 steps). Time spent in MVPA in bouts showed a simiar pattern: both intervention groups increased at 12 months compared with contro participants. Additiona MVPA in bouts (minutes per week) was 33 minutes (95% CI 17 to 49 minutes; p < 0.001,) for the posta group, and 35 minutes (95% CI 19 to Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxvii

30 SCIENTIFIC SUMMARY 51 minutes; p < 0.001) for the nurse-support group, with no statisticay significant difference between intervention groups (2 minutes, 95% CI 14 to 17 minutes). Sedentary time and acceerometer wear time were simiar between groups. The interventions had no significant effects on anthropometric measures, anxiety, depression, heath-reated quaity of ife or pain scores. The 12-month exercise sef-efficacy score was significanty higher in the nurse-support group compared with contro participants. None of the foowing acted as an effect modifier for the intervention effect: age, sex, taking part as a coupe, body mass index, disabiity, pain, socioeconomic group and exercise sef-efficacy. Tota adverse events (sef-reported or from primary care records) and serious adverse events (from tria safety monitoring) were simiar between groups. Economic evauation The incrementa cost per step was 0.19 and 3.61 per minute in a 10-minute MVPA bout for the nurse-support group, whereas the posta group took more steps and cost ess than contro participants. The posta group had a 50% chance of being cost-effective at a 20,000 per QALY threshod within 1 year, and had both ower costs ( 11M, 95% CI 12 to 10) and higher QALYs (759 QALYs gained, 95% CI 400 to 1247 QALYs) than the nurse-support and contro groups in the ong term, with an incrementa net benefit of 26M per 100,000 popuation. Sensitivity anayses argey supported findings, except in the tria anaysis, in which four aternative assumptions were made: (1) extending the perspective to participants, (2) excuding heath service use, (3) using sef-reported adverse events and (4) using 3-month outcome data, when contro dominated posta. Long-term cost-effectiveness resuts were very robust. Process evauation A tota of 256 out of 346 participants (74%) in the nurse-support group attended a three sessions, and 268 out of 339 participants (79%) in the posta group and 281 out of 346 participants (81%) in the nurse-support group returned competed step count diaries. Positive associations were seen between increases in step count and time in MVPA in bouts and between both the number of nurse sessions attended and competed step count diary return. Quaitative evauation Forty-three tria participants were interviewed. The intervention was acceptabe and primary care was an appropriate setting. Amost a participants fet that they had benefited, irrespective of their step count change. Important faciitators incuded a desire for a heathy ifestye, improved physica heath, enjoying waking, having a fexibe routine, appropriate externa monitoring and sef-monitoring and socia support. Important barriers incuded heath probems, an infexibe routine, the weather, work and other commitments. Athough the posta group participants were mainy confident to increase their PA without individuay taiored nurse support, two important caveats were heath probems and overcoming barriers. Practice nurses enjoyed deivering the Pedometer And Consutation Evauation-UP (PACE-UP) intervention, and beieved that taking part, especiay in the BCT training, enhanced the quaity and deivery of support provided within routine consutations. Three-year foow-up Of the 1023 tria participants, 681 (67%) provided adequate acceerometry outcome data. The nurse-support and posta intervention groups both showed persistent effects on the 3-year foow-up PA measures, with no difference between them; for the nurse-support group and the posta group versus the contro group, additiona steps per day were 648 steps (95% CI 272 to 1024 steps), and additiona MVPA in 10-minute bouts (minutes per week) were 26 minutes (95% CI 8 to 44 minutes). Quaitative interview findings at 3 years on factors affecting PA maintenance with intervention group participants compemented earier quaitative findings. The pedometer was reported as kick-starting reguar activity and heping to maintain activity. Factors that faciitated PA eve maintenance were striving to maintain good heath, sef-motivation, socia support and good weather. Lack of time was the most frequenty cited barrier; other barriers were often the reverse of faciitators, such as poor heath or bad weather. Findings from the contro group xxviii NIHR Journas Library

31 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 participants, who were sent the pedometer and materias at 12 months, suggested that many had not used them. The persistent 3-year intervention effects, despite contro participants receiving intervention materias at 12 months, suggest that other posta group factors were important (e.g. teephone contact after sending out materias and returning competed PA diaries). The posta group seemed to require this additiona minima support (not provided face to face, or by a heath professiona) in order to be effective. Concusions The PACE-UP pedometer-based waking intervention increased step counts by about one-tenth, and time in MVPA in bouts by about one-third, at 1 year, in predominanty inactive 45- to 75-year-od primary care patients. Nurse and posta deivery had simiar effects on 12-month PA outcomes. The intervention was safe and acceptabe to patients and nurses. The posta group had a 50% chance of being cost-effective at a 20,000 per QALY threshod within 1 year, and was significanty more cost-effective than nurse support and the contro group in the ong term, thus providing a cost-effective way of deivering ong-term quaity-of-ife benefits. Both intervention groups had persistent positive effects on objective PA eves at 3 years, suggesting ong-term benefit. Impications for heath care A primary care pedometer-based waking intervention, deivered by post with minima support, coud provide an effective and cost-effective approach to addressing the pubic heath physica inactivity chaenge. The 3000 steps in 30 minutes neaty captures intensity and coud become a usefu new pubic heath goa, particuary as many peope can measure steps easiy with their mobie phones. The PACE-UP 12-week pedometer-based waking intervention coud be considered for incusion into the NHS Heath Check programme, aimed at a simiar age group (of 40- to 74-year-ods) and/or the NHS Diabetes Prevention Programme. Recommendations for research The PACE-UP tria generaisabiity is imited by the 10% overa recruitment rate and ower recruitment in Asian and socioeconomicay deprived patients. Further research into different recruitment methods is needed, as is research assessing the recruitment achievabe if this programme were to be offered outside a tria setting over a more proonged time period. Athough overa posta outcomes were as effective as, and more cost-effective than, nurse outcomes, further research is required to understand who woud benefit most from the individua taioring offered by a nurse-supported intervention. There has been a recent dramatic increase in the use of wearabes to monitor persona PA eves, incuding smartphones, wrist-worn devices, onine monitoring and mobie apps. Further research into how the PACE-UP 12-week PA programme coud be integrated into the use of these devices (with or without a pedometer) is needed. Tria registration This tria is registered as ISRCTN Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxix

32 SCIENTIFIC SUMMARY Funding Funding for this study was provided by the Heath Technoogy Assessment programme of the Nationa Institute for Heath Research. xxx NIHR Journas Library

33 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 1 Introduction: why this study was needed Benefits and risks of physica activity and current physica activity guideines What are the benefits of physica activity for aduts and oder aduts? Physica activity (PA) eads to reduced mortaity, a reduced risk of > 20 diseases and conditions, and improved function, quaity of ife and emotiona we-being. 1 Physica inactivity is the fourth eading risk factor for goba mortaity; 2 in 2008, it was estimated to have caused 9% of premature mortaity and 5.3 miion deaths wordwide. 3 Physica inactivity is aso a major cost burden on heath services. 1,4,5 What are the current physica activity guideines and who is achieving them? Aduts and oder aduts are advised to be active daiy and, for heath benefits, shoud achieve at east 150 minutes (2.5 hours) weeky of at east moderate-intensity activity [moderate to vigorous PA (MVPA)] or 75 minutes of vigorous-intensity PA, or an equivaent combination, achieved in bouts of at east 10 minutes duration. 1,2 Musce-strengthening activities are aso recommended on at east 2 days weeky, 1,2 but are not part of our intervention, which is focused on increasing waking. One effective way to achieve the aerobic PA recommendations is by undertaking 30 minutes of moderate-intensity PA on at east 5 days weeky. 1,6,7 Reguar waking is the most common PA of aduts and oder aduts, and waking at a moderate pace (3 mies or 5 km per hour) quaifies as moderate-intensity PA. 8 Time spent being sedentary for extended periods shoud aso be minimised, as this is an independent disease risk factor, 1 which increases steepy with age from 45 years. 9 There is an increasing awareness that as a dose response reationship exists for PA and heath benefits, getting inactive peope to do a itte more PA is aso important, rather than just reying on trying to achieve PA recommendations. 10,11 Emphasising that the 30 minutes can be buit up from 10-minute bouts is an important message for oder aduts and those with disabiities, enabing them to increase their MVPA graduay. Among aduts in Engand aged 19 years, 66% of men and 56% of women sef-reported meeting the recommended PA eves, whereas ony 58% of men and 52% of women aged years did so. 12 Lower socioeconomic groups 9 and Indian, Pakistani, Bangadeshi and Chinese ethnic groups are significanty ess ikey to report meeting the recommended PA eves, whereas the activity eves of other ethnic groups (back Caribbean, back African and Irish) are simiar to that of the genera popuation. 13 Over one-third of aduts wordwide are insufficienty active, but there is arge geographica variation. 10,14 However, PA, incuding waking, is very unreiaby recaed, so surveys overestimate PA eves. 15 Objective acceerometry found that ony 5% of men and 4% of women aged years and 5% of men and 0% of women aged 65 years achieved the recommended PA eves, which is a fraction of those who sef-report achieving them. 9 What are the risks from increasing physica activity? The risks from a sedentary ifestye far exceed the risks from reguar PA. 6,16,17 Moderate-intensity PA carries a ow injury risk, 18 mainy muscuoskeeta injury or fas. 19 Waking is very ow risk, a near perfect exercise. 8 Screening participants for contraindications before participating in ight- to moderate-intensity PA programmes is no onger advocated. 6,20 An important safety feature of our study is that individuaised goas can be set from the participant s own baseine, in ine with the advice that oder aduts in particuar shoud start with ow-intensity PA and increase the intensity graduay: the start ow and go sow approach. 16,17 This worked we with our previous PA tria in oder aduts, which empoyed a simiar approach and showed no increase in adverse events (AEs). 21 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 1

34 INTRODUCTION: WHY THIS STUDY WAS NEEDED Strategies for increasing physica activity How can aduts and oder aduts increase their physica activity eves? A systematic review of PA interventions reported moderatey positive short-term effects, but the findings were imited by mainy unreiabe sef-report measures in motivated vounteers. 22 This review has recenty been updated by three compementary Cochrane reviews assessing (1) face-to-face PA interventions, 23 (2) remote (incuding posta and teephone) and web 2.0 interventions 24 and (3) a direct comparison of these two approaches. 23 Evidence supports the effectiveness of both face-to-face interventions and remote or web 2.0 interventions for promoting PA. However, the reviewers caed for future studies to provide greater detai of the components of face-to-face interventions and to assess the impact on quaity of ife, AEs and economic data, 23 and to incude participants from varying socioeconomic and ethnic groups. 24 Ony one study 25 met the review criteria to compare face-to-face interventions with remote or web 2.0 interventions (many trias were excuded as a resut of having ess than 1 year s foow-up data or an inadequate contro group); this study showed no effect on cardiorespiratory fitness 25 and there were no reported data for PA, quaity of ife or cost-effectiveness. 23 The review concuded that there was insufficient evidence to assess whether face-to-face interventions or remote and web 2.0 approaches are more effective at promoting PA, and caed for more high-quaity comparative studies. 23 None of the studies incuded in the reviews provided objective PA measurement. 23,24 Other studies have concuded that exercise programmes in diverse popuations can promote short- to medium-term increases in PA when interventions are based on heath behaviour theoretica constructs, individuay taiored with personaised activity goas and using behavioura strategies. 26,27 A critica review and a best-practice statement on oder peope s PA interventions advised home-based rather than gym-based programmes, and behavioura strategies (e.g. goa-setting, sef-monitoring, sef-efficacy, support, reapse prevention training), rather than heath education aone. 17,27 Nationa Institute for Heath and Care Exceence (NICE) s guidance concuded that no particuar behaviour change mode was superior and that training shoud focus on generic competencies and skis, rather than on specific modes. 28 More recent compementary NICE guidance specificay recommended that goas, panning, feedback and monitoring techniques shoud be incuded in behaviour change interventions. 29 Starting ow, but graduay increasing to moderate intensity is promoted as best practice, with advice to incorporate interventions into the daiy routine (e.g. waking). 17 A recent systematic review of waking interventions concuded that interventions taiored to peope s needs, targeted at the most sedentary peope and deivered at the eve of the individua or househod, can be effective, athough evidence directy comparing interventions targeted at individuas, coupes or househods was acking. 30 Are pedometers hepfu? Pedometers are sma, cheap devices, worn at the hip, providing direct step count feedback. A systematic review (of 26 studies) found that pedometers increased steps per day by 2491 steps [95% confidence interva (CI) 1098 to 3885 steps] and PA eves by 27%, with significant reductions in body mass index (BMI) and bood pressure. 31 A second review (of 32 studies) found an average increase of 2000 steps per day. 32 Step count goas and diaries were key factors. 31,32 Severa imitations were recognised: study sizes were sma and ong-term effects were undetermined; many studies incuded severa components (e.g. pedometer and support), so independent effects were difficut to estabish; and the incusion of oder peope and men was imited. 31,32 Recent studies have addressed some of these imitations. A pedometer pus behaviour change intervention increased PA at 3 months, but not at 6 months, in 210 oder women, with pedometers providing no additiona benefit. 33 Two trias in high-risk groups showed sustained step count increases at 12 months. 34,35 A recent study of 298 oder aduts found a significant increase in both step counts and time in MVPA at 3 months and 12 months from a practice nurse-deivered pedometer-based waking intervention, but did not separate out pedometer- and nurse support-reated effects. 21 NICE recenty updated its advice from advising the use of pedometers ony as part of research 36 to advising their use as part of packages, incuding support to set reaistic goas, monitoring and feedback NIHR Journas Library

35 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 How do step count goas reate to physica activity recommendations? Step count goas ead to more effective interventions, but no specific approach to goa-setting is favoured. 28 Goas are based on a fixed target (e.g. 10,000 steps per day) 38,39 or on advising incrementa increases from the baseine as a percentage (5% per week, 40 10% biweeky 41 or 20% monthy 33 ) or on a fixed number of extra steps. Those advocating a fixed number of extra daiy steps have deveoped step-based guideines to fit with existing evidence-based guideines with an emphasis on 30 minutes of MVPA on 5 days weeky. 42 Despite individua variation, moderate-intensity waking appears to be approximatey equa to at east 100 steps per minute. 42,43 Mutipied by 30 minutes, this produces a minimum of 3000 steps per day, to be done over and above habitua activity, which is the 3000 in 30 message. 43 Severa studies have advocated adding in 3000 steps per day on most days weeky, either from the beginning 34 or by increasing incrementay (initiay an extra 1500 steps per day and increasing), 44,45 or by increasing by 500 steps per day biweeky. 35 Studies that advised adding 3000 steps per day from the baseine produced significant improvements in step counts at 3 months and two measured outcomes at 12 months, and showed sustained improvements in step counts, 34,35 waist circumference 34 and fasting gucose eves, 35 but no sustained improvements to date in MVPA eves. Athough there is no evidence at present to inform a moderate-intensity cadence (steps per minute) in oder aduts, Tudor-Locke et a. 46 advocate using the adut cadence of 100 steps per minute in oder aduts (whie recognising that this may be unobtainabe for some individuas) and advising that the 30 minutes can be broken down into bouts of at east 10 minutes. This mode was used in a primary care waking intervention in 41 oder peope, which found significant step count increases from baseine to week 12, which were maintained at week ,48 Coud acceerometers be usefu in a pedometer-based waking intervention? Acceerometers are sma activity monitors worn ike pedometers, but are more expensive; however, they are abe to provide a time-stamped record of PA frequency (step counts) and intensity (activity counts). They require computer anaysis, function as binded pedometers in objectivey measuring baseine and outcome data, and provide objective data on time spent in different PA intensities, incuding time spent in MVPA and time spent being sedentary, two important pubic heath outcomes. Pedometer studies without acceerometers have reied on sef-reported measures of these outcomes. Acceerometers are vaid and acceptabe to aduts 9,49 and oder aduts. 9,21,50 53 Athough both instruments measure step count and are highy correated, 50 pedometers usuay record ower step counts, and acceerometers cannot reiaby be substituted for pedometers at an individua eve. 54 Thus, athough we used the acceerometer to measure outcomes, incuding step count, MVPA and sedentary time, we used the binded pedometer, worn simutaneousy at baseine, to set individua step count targets. Are pedometers cost-effective? There is imited knowedge on the cost-effectiveness of pedometer-based interventions in the UK. Recent systematic reviews that considered the economic outcomes of pedometer-based interventions found no evidence, 55,56 party because of an insufficient number of data. 57 However, a recent study assessed the cost-effectiveness of giving an individuaised waking programme and pedometer with or without a PA consutation aongside a community-based tria of 79 peope. 58 The incrementa cost-effectiveness ratios (ICERs) per persons achieving an additiona 15,000 steps per week were 591 and 92 with and without the consutation, respectivey. However, even with this highy seected sampe, no data on quaity of ife were coected, and the impacts on ong-term outcomes were not estimated. What is the roe of primary care in promoting physica activity? Primary care centres (genera practices) in the UK provide heath care and heath promotion, free at the point of access, to a registered ist of oca patients (for many of whom PA wi be of benefit), using disease registers to provide annua or more frequent chronic disease reviews via a mutidiscipinary heath-care team providing continuity of care. NICE guidance found that brief interventions in primary care are cost-effective, and it therefore recommends that a primary care practitioners shoud take the opportunity to identify inactive aduts and provide advice on increasing PA eves. 36 New 5-yeary NHS heath checks incude aduts aged years and incorporate advice on increasing PA, often from primary care nurses. 59 Primary care nurses are effective at increasing PA, particuary waking, in this age group. 60 Not ony can PA Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 3

36 INTRODUCTION: WHY THIS STUDY WAS NEEDED advice through consutation with heath professionas be individuay taiored 61 and have more impact than other PA advice, 62 but this is particuary the case for oder aduts, 63 especiay given the uncertainty about the effectiveness of exercise referra schemes from primary care. 64 Exercise-prescribing guidance in primary care reinforces the importance of foow-up to chart progress, set goas, sove probems, and identify and use socia support; 65 this wi be an important feature of the nurse PA consutations in this tria. Evauation of the UK Step-O-Metre programme, deivering pedometers through primary care, showed sef-reported PA increases, but advised investigation with a randomised controed tria (RCT) design. 44 Two trias, both in oder primary care patients, have assessed the effectiveness of pedometers pus primary care PA consutations; one sma tria (n = 41) showed a significant effect on step counts at 3 months, which was maintained at 6months, 47,48 and the other was our recent Pedometer Acceerometer Evauation-LIFT (PACE-LIFT) tria 21 (n = 298), which showed differences in both step counts and time in MVPA in bouts of at east 10 minutes, at 3 and 12 months for the nurse intervention compared with the contro group. Neither tria separated out pedometer effects from the support provided. 21,47,48 Theoretica base, pioting and preparatory work to deveop the intervention The pedometer-based intervention is based on work cited above, 31,32 showing that pedometers can increase step counts and PA intensity. 31,32 It extends current understanding by aso incuding oder aduts and men, having a 12-month foow-up and ensuring that the pedometer and support components coud be evauated separatey. The patient handbook, diary (both avaiabe on the journas ibrary website: and practice nurse PA consutations use behaviour change techniques (BCTs; e.g. goa-setting, sef-monitoring, feedback, boosting motivation, encouraging socia support, addressing barriers or reapse anticipation). These techniques have been successfuy used by non-speciaists in primary care after brief training, 66 and are emphasised in the Improving Heath: Changing Behaviour: NHS Heath Trainer Handbook, 67 based on evidence from a range of psychoogica methods and intended for NHS behaviour change programmes, with oca adaptation. 67 They aso incude techniques specificay recommended to be incuded in more recent NICE guidance (goas, panning, feedback and monitoring). 29 We adapted the Improving Heath: Changing Behaviour: NHS Heath Trainer Handbook 67 for use in this tria into Pedometer And Consutation Evauation-UP (PACE-UP) nurse and patient handbooks, to focus specificay on PA using pedometers. The BCTs were cassified in accordance with the refined taxonomy of BCTs for PA interventions by Michie et a. 68 Diary recording of pedometer step counts provides cear materia for PA goa-setting, sef-monitoring and feedback, and shoud fit we with this approach. We have adopted the approach used by others 44,45 of advocating adding in 3000 steps per day to an individua s baseine on most days weeky in an incrementa manner, and of advising on graduay increasing PA intensity to achieve more time in MVPA, with the message that 3000 steps in 30 minutes wi hep peope to achieve PA guideines. 43 Reevant piot and preparatory work incudes observationa work using pedometers and acceerometers in primary care 53 and a successfu tria with oder primary care patients deveoping the PA consutations and pedometer-based waking intervention (the PACE-LIFT tria; ISRCTN ,69 ). The PACE-LIFT tria demonstrated that taiored support from practice nurse PA consutations combined with a pedometer-based waking programme (pus acceerometer feedback on PA intensity) ed to an increase in both step counts and time in MVPA compared with the contro group at both 3 and 12 months in 60- to 75-year-od primary care patients. The tria was imited in terms of both ethnic and socioeconomic diversity, has not yet pubished on sedentary time or cost-effectiveness and, as mentioned, was unabe to separate out the effects of the pedometer (and acceerometer feedback) from the effects of nurse support. 21 Rationae for research The PACE-UP tria aimed to fi the gaps in the current evidence base by evauating the effect of a pedometer-based waking intervention, with and without additiona nurse PA consutations in a popuation-based, primary care sampe of inactive aduts and oder aduts. The initia tria incuded foow-up to 1 year and aimed to ensure that adequate numbers of men, oder aduts and individuas from diverse socioeconomic and ethnic backgrounds were incuded. It aso enabed the effectiveness of taking part as an 4 NIHR Journas Library

37 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 individua or as a coupe to be estimated. The intervention used step goas and diaries, and the PA consutations and patient handbook were based on BCTs, such as those used in the Improving Heath: Changing Behaviour: NHS Heath Trainer Handbook. 67 To objectivey test the effectiveness of the intervention on important pubic heath outcomes, such as time spent in MVPA and time spent being sedentary, PA outcomes were assessed by acceerometry. Anonymised practice demographic data were avaiabe for a those invited to participate, enabing the investigation of inequaities in tria participation. Quaitative evauations were aso needed to expore the reasons for tria non-participation, the acceptabiity of the intervention to both participants and practice nurses and the barriers to, and faciitators of, the intervention. An economic evauation was performed aongside the tria and was aso used to inform ong-term cost-effectiveness. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 5

38

39 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 2 Methods This chapter is a summary of the fu study protoco for the tria as originay funded, except for the paragraph which describes changes to the pubished protoco. Some of the materia, incuding the tabes, has aready appeared in pubication, 70 and is reproduced here under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution and reproduction in any medium, provided the origina work is propery cited. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, 70 uness otherwise stated. Further funding was ater awarded by the Nationa Institute for Heath Research (NIHR) Heath Technoogy Assessment programme for a 3-year foow-up of the tria cohort, the methods and resuts of which are described in Chapter 8. Study design The PACE-UP waking intervention tria was a pragmatic, three-arm parae-custer tria (randomised by househod to aow individuas and coupes to participate). It was based in primary care with 45- to 75-year-od inactive aduts, with a 12-month foow-up period, and compared the foowing three groups: contro (usua PA); pedometer and written instructions by post (pedometer by post); and pedometer, written instructions and practice nurse individuay taiored PA consutations (pedometer pus nurse support). Study aims and objectives Study aims The main hypotheses to be addressed were as foows: 1. Does a 3-month posta pedometer-based waking intervention increase PA in inactive 45- to 75-year-ods at the 12-month foow-up point? 2. Does providing practice nurse support through dedicated PA consutations provide additiona benefit? The study aso aimed to assess the cost-effectiveness of both interventions, whether or not any factors modified the intervention effects and the effect of the interventions on patient-reported outcomes, anthropometric measures and primary care-recorded AEs. Primary objectives (reating to the primary outcome of step counts) In inactive aduts aged years, the primary objectives were to: confirm that taiored support from practice nurse PA consutations combined with a pedometer-based waking programme can promote an increase in step counts compared with the contro group at 12 months (pedometer pus nurse support vs. contro) determine whether or not the simpe provision by post of pedometers pus written instructions for a pedometer-based waking programme can promote an increase in step counts compared with the contro group at 12 months (pedometer by post vs. contro) estimate the effect of taiored support from practice nurse PA consutations combined with a pedometer-based waking programme compared with the posta pedometer-based waking programme, on step counts at 12 months (pedometer pus nurse support vs. pedometer by post). Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 7

40 METHODS Secondary objectives (reating to secondary outcomes of time in moderate to vigorous physica activity in bouts, sedentary time and cost-effectiveness) In inactive aduts aged years, the secondary objectives were to: confirm that taiored support from practice nurse PA consutations combined with a pedometer-based waking programme can promote an increase in steps counts at 3 months and time spent in MVPA in 10-minute bouts at 3 and 12 months, and a decrease in sedentary time at 3 and 12 months compared with contro (pedometer pus nurse support vs. contro) determine whether or not the simpe provision by post of pedometers pus written instructions for a pedometer-based waking programme can promote an increase in step counts at 3 months, an increase in time spent in MVPA in 10-minute bouts at 3 and 12 months and a decrease in sedentary time at 3 and 12 months compared with the contro group (pedometer by post vs. contro) estimate the effect of taiored support from practice nurse PA consutations in addition to the pedometer-based waking programme aone on step counts at 3 months, and time spent in MVPA in 10-minute bouts and sedentary time at 3 and 12 months (pedometer pus nurse support vs. pedometer by post) determine the cost-effectiveness of these aternative approaches to increasing PA eves at both 12 months and from a ifetime perspective from the viewpoint of the NHS and participants (see Chapter 4). Other objectives To determine the effect of the interventions on anthropometric measures (BMI, waist circumference and body fat) at 12 months. To determine the effect of the interventions on patient-reported outcomes (sef-reported PA eves, anxiety and depression score, exercise sef-efficacy, quaity of ife, pain, AEs) and on primary care-recorded AEs at 3 months and 12 months. To determine whether or not age groups (< 60 years vs. 60 years), sex, taking part as a coupe, socioeconomic group, disabiity, pain, BMI and exercise sef-efficacy modify the effect of the intervention on increasing step count at 3 months and 12 months (ethnic group was originay intended to be incuded as an effect modifier, but there was inadequate power for this anaysis because of the ow number of non-white participants; see Changes from the pubished protoco). To compare the age, sex, socioeconomic group and ethnicity of those taking part in the tria with those invited but not participating, and to expore the reasons for not participating (see Chapter 5). To assess the fideity and quaity of the intervention impementation over time, by the evauation of patient diary step count goas and recorded step counts for both intervention groups at the 3-month assessment, and the number and timing of recorded practice nurse contacts for the nurse support group (see Chapter 6). To expore the intervention s acceptabiity to practice nurses and inactive aduts, the reasons for dropout and the durabiity of effects, by quaitative interviews with participants after the 12-month foow-up, and a focus group with the nurses on study competion (see Chapter 7). Practice and participant incusion/excusion criteria Practice incusion criteria Genera practices were recruited through the Primary Care Research Network Greater London. Practices were required to be in the south-west London custer, have a practice ist size of > 9000, give a commitment to participate over the duration of the study, have a practice nurse interested and with time to carry out the PA interventions and tria procedures, and have the avaiabiity of a room for the research assistant to recruit participants and carry out baseine and foow-up assessments. 8 NIHR Journas Library

41 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Participant incusion criteria Participants were patients aged years, who were registered with one of the recruited south-west London genera practices, were abe to wak outside the home and had no contraindications to increasing their MVPA eves. Participant excusion criteria Physica activity based (by screening question on invitation etter). In order to maximise the benefits of the intervention to individuas and the NHS, the tria focused on ess-active aduts, using a singe-item vaidated questionnaire measure of sef-reported PA as a screening question to identify them. 60 Those who reported achieving a minimum of 150 minutes of MVPA weeky 1 on their response etter were excuded (participants who, on subsequent baseine acceerometer assessment, were found to be above this PA eve were not excuded, as they woud be incuded if this intervention were to be roed out in primary care). Heath based [either by the Read code from primary care records or by genera practitioner (GP)/practice nurse opinion, or from the teephone or face-to-face baseine assessment with the research assistant] for the foowing reasons: housebound or iving in a residentia or nursing home three or more fas in the previous year, or one or more fas in the previous year requiring medica attention termina iness dementia or significant cognitive impairment registered bind new-onset chest pain, myocardia infarction, a coronary artery bypass graft or an angiopasty within the ast 3 months a medica or psychiatric condition that the GP (or practice nurse) considered to excude the patient (e.g. acute systemic iness such as pneumonia, acute rheumatoid arthritis, unstabe/acute heart faiure, significant neuroogica disease/impairment, unabe to move about independenty, psychotic iness) pregnancy. Recruitment of practices and participants: informed consent Practice recruitment The Primary Care Research Network Greater London identified practices that fitted the above practice incusion criteria. Practice recruitment was chaenging for a number of reasons, incuding difficuties in finding practices with sufficient space to accommodate a research assistant on a reguar basis, finding practices with nurses wiing and with sufficient time to be engaged in deivering the intervention and finding practices that were prepared to provide administrative support. The Primary Care Research Network Greater London provided us with strong support to recruit practices. Initiay, six practices were recruited, with an additiona practice added haf-way through to boost recruitment. This was necessary, as recruitment at that point was running at just beow 10% and we were concerned that we woud not achieve the target recruitment from the origina six practices within 12 months. The practices were seected to incude a range of sociodemographic factors and geographica circumstances based on the practice postcode Index of Mutipe Deprivation 71 (IMD) scores (at east one practice from each quintie). Participant recruitment Practice staff identified patients aged years on their primary care eectronic patient record system, and, using Read codes and oca care home knowedge, excuded ineigibe patients (patients were aged years when seected, but some were aged 75 years by the time of recruitment or randomisation). Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 9

42 METHODS A ist of potentiay eigibe patients was produced and ordered by househod, with a unique househod identifier number. An anonymised ist was then used by the research team to create at east four random sampes of 400 individuas at each practice. A maximum of two peope per househod were seected (we were aiming to seect coupes). If a househod had two individuas, one was seected at random, and if the second individua had an age difference of 15 years, they were aso seected; if they fe outside this age range, they were not incuded. If a househod had more than two individuas, one was seected at random, and if there was a second individua aged within 15 years they were aso seected; if there was not a second individua aged within 15 years, this became an individua househod. Each sampe ist was examined by practice nurses or GPs to ensure tria suitabiity prior to invitation (see Participant excusion criteria). Participants were recruited between September 2012 and October 2013, and foow-up was competed by October Non-responders and non-participants See Chapter 5 for more detais. Informed consent Patients were sent an invitation etter from their own practice, aong with a participant information sheet and a screening question on sef-reported PA. A reminder invitation was sent if no repy was received after 6 weeks. A og was kept of the response rates from each practice. The decision regarding participation in the study was entirey vountary. Those interested in participating returned the repy sip, incuding a response to a singe screening question about their usua PA eves. If the participant sef-reported as not achieving the PA guideines, 1 the research assistant arranged a baseine appointment for them and ran through the participant information sheet, and handed any questions or concerns that they had. If they were happy to proceed, they signed the study consent form; this form incuded consent to be contacted for quaitative interviews and consent for their genera practice records for the year of the tria to be downoaded after tria foow-up was competed. Participants who had difficuty understanding, speaking or reading Engish were accompanied by a famiy member or friend during the research assistant appointment. Participants within a coupe coud attend together or separatey. Changes from the pubished protoco We panned to recruit from six genera practices, but to enabe target recruitment, a seventh practice was added in December Changes from protoco-panned anayses 70 were approved by the Tria Steering Committee (TSC), prior to anaysis. We report MVPA in 10-minute bouts, as this reates more cosey to PA guideines. 1,2 Ony 20% of participants were non-white; ethnic group was therefore excuded from the subgroup anayses, as a resut of ow power. Interventions Tabe 1 shows the components of the intervention provided to the posta and nurse groups. Tabe 2 shows the content of the patient handbook and the patient diary and the BCTs that were incuded in each of them, rated according to Michie et a. s CALO-RE taxonomy. 68 Tabe 3 shows the timing and session content for the three dedicated nurse PA consutations and the BCTs intended to be covered in each session. Figure 1 provides a summary of the 12-week waking programme in terms of steps per day or time spent waking, to be added to each individua s baseine average daiy steps. The training received by the practice nurses in order to deiver the interventions is described in Chapter 6. A figure summarising the tria procedures and compex intervention components is shown in Appendix 1, Figure NIHR Journas Library

43 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 1 Components of the compex intervention for the PACE-UP tria Component What was provided Tria arm receiving Additiona detais on components Pedometer Yamax Digi-Waker (Yamasa Tokei Keiki Co., Ltd, Tokyo, Japan), SW-200 mode Posta group pedometer posted with instructions Nurse-support group pedometer given with instructions by the nurse Provided direct step count to participants. Required daiy manua recording and resetting PACE-UP handbook, 12-week waking pan and step count diary a Handbook to support the 12-week waking programme. Individuaised waking pan (see Figure 1). Diary to record weeky step count and waks for 12 weeks Posta group posted Nurse-support group given by nurse Baseine average daiy step counts (from the binded pedometer assessment) were used to create individua targets. The 12-week waking programme graduay increased targets to achieve an additiona 3000 steps per day (approximatey 30 minutes of brisk waking) on 5 days weeky. Daiy step counts and target achievement were recorded in the diary. Tabe 2 ists BCTs in the PACE-UP handbook and diary Practice nursededicated PA consutations Three individuay taiored consutations. Participants coud be seen individuay or as a coupe Nurse-support group ony Session timings, content and panned BCTs are shown in Tabe 3. Sessions reinforced the intervention defined in the diary and the handbook. The nurse consutation aowed some additiona BCTs to be used and provided an opportunity to individuay taior the intervention to participants needs a The handbook, 12-week waking pan and step count diary are avaiabe on the NIHR Journas Library website ( This tabe has been adapted from Harris et a. 70 This artice is pubished under icense to BioMed Centra Ltd. This is an Open Access artice distributed under the terms of the Creative Commons Attribution License ( by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the origina work is propery cited. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, uness otherwise stated. Procedure for the posta intervention group The participants received (by post) a pedometer (Yamax Digi-Waker, SW-200 mode; Yamasa Tokei Keiki Co., Ltd, Tokyo, Japan), instructions and a 12-week step count diary for the 12-week waking intervention (see Figure 1). The research assistant contacted the participant to check that the pedometer had been received and to resove any difficuties with the equipment. At the end of the 12-month foow-up period, the posta group were offered a singe practice nurse PA appointment, if they wanted it. Procedure for the nurse intervention group Three dedicated PA consutations (week 1, week 5 and week 9) were arranged with the practice nurse, to individuay taior and support the 12-week pedometer-based waking programme (see Figure 1). At their first appointment, participants were given the same pedometer, diary and handbook that the posta group received. Participants were asked to wear a pedometer and keep a diary record of daiy steps for 4 weeks between appointments, in order to review targets and goas at their next appointment. Participants were seen individuay or as a coupe. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 11

44 METHODS TABLE 2 The PACE-UP patient handbook, and diary and BCTs incuded Component Guide to content a BCTs 68 Patient handbook Heath benefits of increasing waking PA guideines Moderate-intensity PA and reating it to number of steps PACE-UP waking programme and step count targets Review participant baseine step count How to increase PA safey Usefu websites How to keep going when the PACE-UP programme finishes 1 and 2 4 7, 9 and , 1, 2, 26, 29 and 35 Patient diary How to use the pedometer and record steps in diary Frequenty asked questions on the PACE-UP tria Weeky recording of step count and waking in diary (weeks 1 12) Achievement of targets (weeks 1 12) Panning when to wak, where to wak, who to wak with Week 2, tips and motivators: make waking part of your daiy routine Week 3, tips and motivators: remember persona benefits, what to do if you are faing behind your targets Week 4, keep it up: praise and reward yoursef, encouraging socia support Week 5, keep motivated: write down step counts, ask for support Week 6, now we are moving: obstaces and soutions Week 7, how to make these changes permanent ideas for new waks, making time for waking, what gains have been made so far? Week 8, maintain the gain: pacing, tips for safe exercising Week 9, be busy being active: keep monitoring with the pedometer, paces, peope and thoughts that motivate you Week 10, change does not happen in a straight ine! Preparing for setbacks Week 11, make it a heathy habit: buiding reguar exercise habits, creating if-then pans Week 12, I ve changed: how to keep up your waking programme Congratuations, you have competed the programme How to keep going when the PACE-UP programme finishes 16 and 21 7, 9, 19 and 26 10, 12 and and , 20 and 35 12, 13 and 29 12, 16 and , 17 and 11 9, 21 and 35 16, 29 and 36 8 and 35 1, 2, 7 and 23 16, 20 and 29 11, 16 and 17 1, 16 and 29 a 1 provide genera information on behaviour heath ink; 2 provide information on consequences to the individua; 4 provide normative information about others behaviour; 7 action-panning; 8 barrier identification; 9 set graded tasks; 10 prompt review of behavioura goas; 11 prompt review of outcome goas; 12 prompt rewards contingent on effort; 13 prompt rewards contingent on successfu behaviour; 16 prompt sef-monitoring of behaviour; 17 prompting sef-monitoring of behavioura outcome; 19 provide feedback on performance; 20 provide information on when and where to perform the behaviour; 21 provide instructions on how to perform the behaviour; 23 teach to use prompts/cues; 26 prompt practice; 29 pan socia support/socia change; 35 reapse prevention/coping panning; 36 stress management/ emotiona contro training; and 38 time management. This tabe has been adapted from Harris et a. 70 This artice is pubished under icense to BioMed Centra Ltd. This is an Open Access artice distributed under the terms of the Creative Commons Attribution License ( by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the origina work is propery cited. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, uness otherwise stated. Procedure for the contro group The participants were advised to continue their usua activity eves and were not offered the 12-week waking intervention, but were free to participate in any other PA, just as they woud if they were not enroed in the tria. At the end of the 12-month foow-up period, the contro group was offered to receive a pedometer and the PACE-UP 12-week waking programme handbook and diary, either by post or as part of a singe PA practice nurse appointment, as preferred. 12 NIHR Journas Library

45 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 3 The PACE-UP practice nurse PA consutations and BCTs incuded Sessions Guide to session content a BCTs 68 Session1: week 1, first steps (30 minutes) Review heath status, current activity, heath benefits of PA Cost benefit anaysis for increasing PA PA guideines and how to increase PA safey Moderate-intensity PA and reating it to number of steps Review participant baseine step count Teach use of pedometer and recording waks and steps in diary Ideas for increasing steps Goa-setting PACE-UP goas or taiored to the individua patient Use of rewards for effort and for achieving goas Summarise and check patient understanding, pan time for next meeting Communication strategies to overcome resistance and promote patient-ed change 1 and and and , 9 and and Session 2: week 5, continuing the changes (20 minutes) Review step count and waking diary Encourage progress in increasing waking and achieving step count goas Troubeshoot any probems with pedometer or diary Review target and agree goas for next stage Barriers to, and faciitators of, increasing PA, overcoming barriers, encouraging support Pacing and avoiding boom and bust Check confidence eves, buid confidence to make change Summarise and check patient understanding, pan time for next meeting Communication strategies to overcome resistance and promote patient-ed change 10 and and , 9 and 16 8 and 29 9 and 35 18, 29 and Session 3: week 9, buiding asting habits (20 minutes) Review step count and waking diary Review overa progress over the sessions Encourage progress in increasing waking and achieving goas Preparing for setbacks Buiding habits: discuss methods of maintaining asting change, incuding repetition, if-then rues and support Setting goas: maintaining current activity or increasing further? Reminder regarding contact with research assistant in 3 4 weeks Communication strategies to overcome resistance and promote patient-ed change 10 and and and , 23, 29 and 35 7, 9, 16 and a 1 provide genera information on behaviour-heath ink; 2 provide information on consequences to individua; 4 provide normative information about others behaviour; 7 action-panning; 8 barrier identification; 9 set graded tasks; 10 prompt review of behavioura goas; 11 prompt review of outcome goas; 12 prompt rewards contingent on effort; 13 prompt rewards contingent on successfu behaviour; 16 prompt sef-monitoring of behaviour; 17 prompting sef-monitoring of behavioura outcome; 18 prompting focus on past success; 19 provide feedback on performance; 20 provide information on when and where to perform the behaviour; 21 provide instructions on how to perform the behaviour; 23 teach to use prompts/cues; 26 prompt practice; 29 pan socia support/socia change; 35 reapse prevention/coping panning; 36 stress management/emotiona contro training; and 37 motivationa interviewing. This tabe has been adapted from Harris et a. 70 This artice is pubished under icense to BioMed Centra Ltd. This is an Open Access artice distributed under the terms of the Creative Commons Attribution License ( icenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the origina work is propery cited. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, uness otherwise stated. Outcome measures Primary and secondary outcome measures These were seected to refect the needs of the target popuation, heping aduts and oder aduts to increase their PA, particuary through waking, and to inform UK pubic heath poicy. The primary outcome was the change in average daiy step count, measured over 7 days, between baseine and 12 months, assessed Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 13

46 METHODS FIGURE 1 A summary of the PACE-UP waking programme. Adapted from Harris et a. 70 This artice is pubished under icense to BioMed Centra Ltd. This is an Open Access artice distributed under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution, and reproduction in any medium, provided the origina work is propery cited. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, uness otherwise stated. objectivey by acceerometry (GT3X+; ActiGraph LLC, Pensacoa, FL, USA). Secondary outcomes were as foows: changes in step counts between baseine and 3 months; changes in time spent weeky in MVPA in 10-minute bouts between baseine and 3 months, and between baseine and 12 months; and time spent sedentary between baseine and 3 months, and between baseine and 12 months. A of these secondary outcomes were aso assessed objectivey by acceerometry. Cost-effectiveness was aso a secondary outcome in our protoco [incrementa cost per change in step count and per quaity-adjusted ife-year (QALY)]; this is presented in Chapter NIHR Journas Library

47 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Anciary outcomes Change in sef-reported PA, measured over the same 7 days as acceerometry using the short Internationa Physica Activity Questionnaire (IPAQ) 72 and the Genera Practice Physica Activity Questionnaire [(GPPAQ) 73 as part of the 7-day PA questionnaire; see Appendix 1]. Change in other patient-reported outcomes (from the heath and ifestye questionnaires at baseine, 3 months and 12 months; see Appendix 1): confidence in abiity to do PA, as measured by exercise sef-efficacy; 74 anxiety and depression, as measured by the Hospita Anxiety and Depression Scae (HADS); 75 perceived heath status (heath-reated quaity of ife), as measured by the EuroQo-5 Dimensions, five-eve version (EQ-5D-5L); 76 and sef-reported pain, measured by two items from the Medica Outcomes Study 36-item short-form heath survey. 77 Change in anthropometric measurements (weight, BMI, waist circumference, body fat; see Ascertainment of outcomes, Anthropometry, for measurement detais). Adverse outcomes [fas, fractures, injuries, exacerbation of pre-existing conditions, major cardiovascuar events and deaths from serious AEs (SAEs)] were coected as part of safety monitoring for the tria, by questionnaire sef-report items designed by us at 3 and 12 months, and from primary care records after the 12-month foow-up period, for those giving consent. Heath service use for those giving consent to primary care record access for the 12-month tria period, numbers of the foowing occurrences were coected for heath economic evauations (see Chapter 4): primary care consutations, accident and emergency (A&E) attendances, emergency and eective hospita admissions and outpatient referras. Ascertainment of outcomes See Figure 2 for the schedue for outcome assessments and measures. Acceerometry Participants were asked to carry on with their usua PA eves and to wear an acceerometer (GT3X+, ActiGraph LLC) on a bet over one hip, during waking hours (from rising unti going to bed) for 7 days, ony removing it for bathing, at baseine, 3 months and 12 months. Participants were offered the option of text messaging to remind them to wear the acceerometer each day and to return it after the 7 days. A diary was provided to record what activities were done and for how ong. The monitor, bet and diary were posted back on competion. Once returned, the participants received a 10 gift voucher. Anthropometry At the baseine and 12-month face-to-face assessments, the foowing measurements were taken: height (measured in bare feet to the nearest 0.5 cm using a stadiometer), weight (measured to the nearest 0.1 kg), body fat, bioimpedence [using the Tanita body composition anayser BC-418 MA (Tanita Corporation, Tokyo, Japan)] and waist and hip circumference (using a standard technique and tape measure with a cear pastic sider). Questionnaire measures Questionnaire measures were coected using vaidated toos (detaied under Outcome measures, Anciary outcomes), as part of sef-competed questionnaires at 3 months and 12 months. In terms of the sef-reported PA in the previous week, for the IPAQ 72 we used the measure of MVPA (tota minutes of vigorous and moderate PA weeky) and the measure of waking (tota minutes of waking weeky). The GPPAQ 73 provides a PA index (PAI), which is cacuated from a combination of PA from both work and eisure activities. Active individuas are those who sef-report 3 hours of MVPA per week on Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 15

48 METHODS the PAI. However, waking is not incuded in the cacuation of the PAI, athough it is asked about in the questionnaire. Anaysis of simiar GPPAQ data compared with objective acceerometry in our earier PACE- LIFT tria 21 demonstrated that a modified PAI, which aso incuded waking at a brisk pace for at east 3 hours per week, improved vaidity and repeatabiity compared with the standard GPPAQ. 78 A modified index, GPPAQ-Wak, was therefore generated. In addition, the foowing were aso recorded at baseine: demographic information, based on 2011 census questions 79 (e.g. marita status, ethnic group, occupation, empoyment, househod composition, home ownership); a ist of common sef-reported chronic conditions (e.g. heart disease, ung disease, arthritis, stroke, diabetes meitus, depression); disabiity, as measured by the Townsend score; 80 imiting ongstanding iness; 79 current medications; smoking; and acoho consumption. Severa other questionnaire variabes were coected at a three time periods, but were not considered to be anciary tria outcomes in the tria protoco: 70 oneiness, measured by a singe item; 81 risk of fas, measured using the Fas Risk Assessment Too, 82 was assessed using a combination of both sef-reported items and direct observation of the abiity to rise from a chair without using arms; and sef-reported usua PA, as measured by the modified Zutphen Physica Activity Questionnaire. 83 Data from these variabes are not presented in this report. A study groups were asked about fas, injuries, fractures, exacerbation of any pre-existing conditions and the costs of any treatments in the 3- and 12-month questionnaires. Questions on the financia costs of participating in waking and other PAs were asked in the 3- and 12-month questionnaires. Primary care computerised record measures For participants who gave written consent, the foowing data were coected from their eectronic primary care records, for the 12-month duration of the tria, after the 12-month foow-up: adverse events potentiay reating to tria participation [Read codes reating to fas, fractures, injuries, cardiovascuar events (myocardia infarction, coronary artery bypass graft, coronary angiopasty, transient ischaemic attack, stroke) and death] heath service use GP consutations, practice nurse consutations (excuding those for the tria), A&E attendances, emergency and eective hospita admissions and outpatient referras. These data were downoaded and pseudoanonymised before remova from the practice. Baseine and foow-up data coection Baseine data coection At baseine, a face-to-face assessment with the research assistant occurred at the participant s genera practice, and questionnaire and anthropometric data were coected (see Ascertainment of outcomes). Participants were then given a bet with an acceerometer (GT3X+, ActiGraph LLC) and a binded pedometer (Yamax Digiwaker CW200) on it, and asked to wear this for 7 days. The CW200 pedometer mode was used to enabe the baseine target-setting of the pedometer step count, because of its 7-day memory of consecutive daiy steps. However, it is buky to wear and compicated to use, so this mode was not used for the intervention. Foow-up data coection Foow-up data coection was conducted in the same way for a tria groups (Figure 2): (1) 3 months (posta) after randomisation (questionnaires and acceerometry) and (2) 12 months (face to face) after the baseine assessment (questionnaires, acceerometry and anthropometry). Participants were aso contacted 16 NIHR Journas Library

49 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 by the research assistant at 6 and 9 months after randomisation by teephone or e-mai to check on fas for tria safety reporting and contact detais. For those in the intervention groups, a repacement pedometer or batteries were offered at each contact point, if required. The intervention groups were asked to return their 12-week step count diary foowing the intervention at 3 months. This was then photocopied and sent back to participants. Acceerometer data reduction The acceerometer measured vertica acceerations in magnitudes from 0.05 to 2.0 g samped at 30 Hz, then summed over a 5-second epoch time period. ActiGraph data were reduced using Actiife software (v 6.6.0; ActiGraph LLC, Pensacoa, FL, USA), ignoring runs of 60 minutes of zero counts. 70 Vertica counts were used, as these are the basis of the vaidated step count and MVPA agorithms. The anaysis summary variabes used were step counts, acceerometer wear time, time spent in MVPA ( 1952 counts per minute, equivaent to 3 metaboic equivaents), 84 time spent in 10-minute MVPA bouts and time spent being sedentary ( 100 counts per minute, equivaent to 1.5 metaboic equivaents). 85 Adverse events and serious adverse events An AE was defined as any unfavourabe and unintended sign, symptom, syndrome or iness that deveoped or worsened during the observation period of the tria. This incuded: exacerbation of a pre-existing iness an increase in frequency or intensity of a pre-existing condition a condition detected or diagnosed after the tria started (but which might have been present at baseine) a persistent disease or symptoms present at baseine that worsened foowing the start of the tria. Baseine 3 months 12 months Type of contact with research assistant Face to face at practice and post Teephone and post Face to face at practice and post Seven-day objective PA acceerometer assessment Questionnaire heath and ifestye measures (incuding the EQ-5D-5L, sef-efficacy, pain, anxiety, depression) Questionnaire 7-day PA reca (sef-report: IPAQ short version and GPPAQ) Anthropometric measures (weight, BMI, waist circumference and body fat percentage) FIGURE 2 Schedue of outcome assessment measures used in the PACE-UP tria. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 17

50 METHODS A SAE was defined as any AE occurring during the tria for any of the three groups that resuted in any of the foowing outcomes: death a ife-threatening AE inpatient hospitaisation or proongation of existing hospitaisation a new disabiity/incapacity. A AEs were assessed for seriousness, expectedness and causaity. A AEs were recorded and cosey monitored unti resoution or stabiisation, or unti it had been shown that the study intervention was not the cause. Participants were asked to contact the tria site immediatey in the event of any SAE. The chief investigator was informed immediatey and determined seriousness and causaity in conjunction with two other medicay trained tria investigators. A SAE that was determined to be directy or possiby tria reated was reported, within agreed time frames, to the TSC and the ethics committee. A SAEs were reported annuay to the TSC, ethics committee and the tria sponsor. Athough it was important to record AEs contemporaneousy for tria safety monitoring during the tria, there was a risk of bias in their reporting, with those having nurse contact having more opportunities for reporting fas, injuries and inesses. For anayses and reporting, we therefore concentrated on measures for which there were fewer risks of bias between groups: (1) spontaneousy reported SAEs, (2) fas, fractures and injuries from questionnaire sef-report at 3 and 12 months and (3) fas, fractures, injuries, cardiovascuar events (new episode of any of the foowing: myocardia infarction, angiopasty, coronary artery bypass, stroke, transient ischaemic attack, new-onset angina, ischaemic heart disease) and deaths from primary care records after the 12-month foow-up point. Sampe size A tota of 217 patients in each of the three tria arms woud aow a difference of 1000 steps per day to be detected between any two arms of the tria, with 90% power at the 1% significance eve. However, we panned to randomise househods. Assuming an intracuster correation of 0.5 and an average househod size of 1.6 eigibe patients, we needed to anayse 282 patients per tria arm. Aowing for approximatey 15% attrition, we needed to randomise a tota of 993 patients (331 contro participants, 331 participants receiving a pedometer by post and 331 participants receiving a pedometer pus nurse support). We initiay panned on six practices to each recruit approximatey 166 patients (approximatey 55 participants to each of the three groups), but to enabe target recruitment, a seventh practice was added. We anticipated a 20% recruitment rate among eigibe participants, based on other PA interventions (incuding with pedometers) among midde-aged and oder aduts in primary care, where the recruitment rate was between 17% and 35%. 21,33,86 89 We estimated that, even if our recruitment rate was as ow as 10%, we woud have enough eigibe participants at practices. In fact, the recruitment rate dipped to beow 10%, so a seventh practice was added. Randomisation, conceament of aocation, contamination and treatment masking Randomisation and conceament of aocation Foowing competion of the baseine assessment (incuding providing acceerometry data on 5 compete days of 9 hours/540 minutes), each participant was aocated to a tria group using the King s Coege London cinica trias unit internet randomisation service, to ensure independence of the aocation. If participants were unabe to provide at east 5 days of 540 minutes wear time on acceerometry, they were asked to wear the acceerometer for a further 7 days, or they were excuded if this was not possibe. 18 NIHR Journas Library

51 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Randomisation was at a househod eve. Randomisation of a group househod took pace ony after both members of the househod had competed the baseine assessment. Bock randomisation was used within the practice with randomy sized bocks (2, 4 or 6) to ensure baance in the groups and an even nurse workoad. Participants were informed by teephone which group they had been aocated to. Contamination Contamination coud occur between partners in a househod; we minimised this by ensuring that, if two househod members were recruited, they were aocated to the same group (i.e. randomisation was at a househod eve). Contamination woud have occurred if the contro group used a pedometer to increase their waking during the 12-month tria foow-up. We tried to discourage participants in the contro group from buying a pedometer, by ensuring that they knew that they woud receive one at the end of foow-up. A question was incuded in the 12-month questionnaire to ask if they had used a pedometer during the course of the tria. Treatment masking Participants were randomised ony after the successfu return of acceerometers with 5 days recording. It was not possibe to mask participants to their intervention group. The research assistants who carried out foow-up assessments were not masked to group aocation for pragmatic reasons aone: the study was funded to support ony enough researchers to carry out recruitment and foow-up simutaneousy. However, the main outcome was assessed objectivey through acceerometry, and the assessment of the quaity of the outcome data was done bind to intervention group: days with < 9 hours of data were excuded. Weight and body fat were aso assessed objectivey using the Tanita scaes, which provided eectronic printouts of resuts, and other outcomes were assessed using standardised measures (e.g. patient-reported outcomes from questionnaires). The statistician carrying out the primary anayses was masked to group aocation as far as possibe. Withdrawas, osses to foow-up and missing data Withdrawas and osses to foow-up Participants coud withdraw from the tria at any point. Participants who withdrew foowing informed consent, and prior to randomisation, were repaced with another participant. Participants who withdrew after randomisation were not repaced and were asked if they were prepared to contribute to further data coection on outcomes at 3 and 12 months. Participants were made aware that withdrawa from the tria woud not affect future care and that information on those who withdrew or were ost to foow-up that had aready been coected woud sti be used, uness consent for this was withdrawn. Procedure for accounting for missing data Ony days with at east 540 minutes of registered time on an acceerometer on a given day were used, which was consistent with previous work (the PACE-LIFT tria, 21 Trost et a. 90 and Mier et a. 91 ). Participants were randomised ony if they provided at east 5 such days of acceerometer data at baseine. A mutieve inear regression mode was used, taking account of repeated days within individuas to estimate the baseine average daiy step count for each subject, adjusted for the day of the week and the day order of wearing the acceerometer. The same approach was used to estimate the average daiy step count at 3 months and 12 months. The main covariates age, sex, practice, month of baseine acceerometry and whether or not participants were taking part as a coupe were known for a participants, and most patients had compete data for other measures. To essen attrition bias, the primary anaysis incuded a participants with at east 1 satisfactory day of acceerometry recording at 12 months (i.e. a wear time of 540 minutes). The main anaysis assumed that, depending on the mode covariates, outcome data were missing at random. This was ikey to be true for missing data as a resut of acceerometer faiure, and was pausibe for missing days and participants who did not return acceerometers. However, an aternative pausibe assumption is that participants who faied to provide outcome data were ess active. Mutipe imputation was used to impute vaues for those with no acceerometer data at 12 months (see Statistica Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 19

52 METHODS methods, Sensitivity anayses). Further sensitivity anayses examined the impact of assuming that missing step counts at 12 months in the contro group were equa to their baseine vaues and, in the two intervention groups, varied between 1500 steps ower and 1500 steps higher than their baseine vaues. Statistica methods The anaysis and reporting was in ine with the Consoidated Standards of Reporting Trias (CONSORT) guideines, with the primary anaysis on an intention-to-treat basis. That is, a participants with outcome data were incuded, regardess of their adherence to the interventions. A participants were incuded in the primary anaysis if they had at east 1 satisfactory day of acceerometer recording ( 540 minutes) of registered time during a day, out of 7 days, at 12 months. The adequacy of the randomisation process to achieve baanced groups was checked by comparing participant characteristics in the three arms (e.g. sex, age, socioeconomic group, baseine PA eve or BMI). Stata, version 12 (StataCorp LP, Coege Station, TX, USA) was used for a anayses. Primary anaysis The primary outcome measure was change in average daiy step count from baseine to the 12-month foow-up point measured over 7 days. However, to overcome Lord s paradox, 92 the anaytic approach regressed 12-month outcomes on baseine measures, thus aowing for regression to the mean. Eigibiity was defined on the basis of 5 days with 540 minutes activity at baseine. If the participant was asked to wear the acceerometer for a second time, the second 7 days was used in the anaysis. If there were > 7 days wear on the acceerometer, then the first 7 days were used and ater readings were discarded. The primary anaysis used a participants providing at east 1 day of 540 minutes acceerometry wear time at 12 months (i.e. a compete-case anaysis). A anayses were carried out using Stata, version 12. The xtmixed procedure was used for regression modes. A two-stage process was used for acceerometry data. Stage 1 estimated the average daiy step count at both baseine and 12 months, using a mutieve mode in which daiy step counts were regressed on day of the week and day order of wearing the acceerometer (from day 1 to day 7) as fixed effects, and with day within individua as the random effect (i.e. eve 1 was the day within individua and eve 2 was individua). In stage 2, average daiy step count at 12 months was regressed on baseine average daiy step count, sex, age, genera practice, month of baseine acceerometry and treatment group as fixed effects, and househod as the random effect, to aow for custering at a househod eve (i.e. eve 1 was individua and eve 2 was househod). This method effectivey measured the change in step count from baseine to 12 months, minimising bias and maintaining power. Adjusting for baseine steps controed for many factors that predict the number of steps in cross-sectiona anayses (e.g. BMI, socioeconomic group, heath status). The reference group for the intervention group comparisons was the contro group. The post-estimation command pwcompare was used to obtain the estimates of change with 95% CIs and p-vaues for the difference in change in steps for the posta group versus the contro group, the nursesupport group versus the contro group and the nurse-support group versus the posta group. This ast comparison provided information on whether or not the nurse intervention promoted a worthwhie increase in activity compared with a pedometer aone. It shoud be noted that, athough this estimate can be obtained from the difference of the first two estimates, pwcompare aso provided 95% CIs for this comparison. Checks were carried out to confirm that the distribution of residuas from the regression mode for change in steps was normay distributed. Secondary and anciary outcome anayses Secondary PA outcome measures from acceerometry were tota weeky minutes of MVPA in 10-minute bouts, average daiy sedentary time at 12 months and steps, MVPA in bouts and sedentary time at 3 months. These data were processed and anaysed in the same way as described for the step counts (see Primary anaysis). MVPA was highy positivey correated with step counts and sedentary time was negativey correated with step counts. 20 NIHR Journas Library

53 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Other anciary outcomes were changes in exercise sef-efficacy, anxiety, depression, perceived heath status (heath-reated quaity of ife, as measured using the EQ-5D-5L), sef-reported pain, anthropometric measures (weight, BMI, waist circumference, body fat) and sef-reported PA from the IPPAQ and GPAQ questionnaires. Changes in these outcomes from baseine to 3 and 12 months were anaysed using identica modes to stage 2, as described in Primary anaysis (i.e. eve 1 was individua and eve 2 was househod). Adverse event anayses The number of participants who suffered an AE between 0 and 3 months or between 0 and 12 months (a spontaneousy reported SAE or a systematicay reported SAE from the 3- or 12-month questionnaire, or a SAE coected from the primary care record data) was compared between groups using exact tests for categorica tabes. Subgroup anayses Sex, age groups (< 60 years or 60 years), taking part as a coupe, socioeconomic subgroups, BMI, disabiity, pain and exercise sef-efficacy were examined as potentia effect modifiers by adding interaction terms to the regression mode for the primary outcome, which was changes in step counts at 12 months. Sensitivity anayses Sensitivity anayses were carried out for the primary outcome (change in step counts from baseine to 12 months). The effects of using different criteria for defining satisfactory wear at 12 months were examined as foows: (1) at east 5 days of 540 minutes wear time, (2) 1 day of 600 minutes wear time and (3) 5 days of 600 minutes wear time. The effect of adjusting for change in wear time between baseine and 12 months was aso examined. Additiona sensitivity anayses assessed whether participants ost to foow-up or who faied to provide a singe adequate day s recording might have introduced bias. This was first done by assuming that outcome data were missing at random, depending on the mode covariates, using the Stata procedure mi impute. The first mode used the standard mode covariates to impute missing step counts at 12 months (treatment group, baseine steps, sex, age, genera practice, month of baseine acceerometry and househod as a random effect) and the second mode added in the Nationa Statistics Socio-economic Cassification (NS-SEC), sef-reported pain and fat mass as additiona covariates. Further anayses expored the possibe impact of outcomes not being missing at random, using the foowing assumptions: among those with missing data in the contro group, the change in mean steps from baseine to 12 months was 0, and among those with missing data in each of the intervention groups, the change in mean steps from baseine to 12 months was 1500, 0 or Ethics approva and research governance Ethics approva was granted for the tria from London, Hampstead Research Ethics Committee (reference number 12/LO/0219). The NHS Research and Deveopment approva was granted by the Cinica Commissioning Groups in south-west London, through the Primary Care Research Network, to cover a the practice sites. Management of the tria The tria progress, incuding recruitment, safety, finance and data management, was reviewed reguary by the tria management group (TMG). This was made up of the chief investigator, two tria investigators, the tria statistician and the tria manager. The TMG met on a monthy basis. A of the tria investigators met as a group (the tria investigator group) on a biannua basis, and the TSC met prior to participant recruitment, and then annuay or biannuay as necessary. Minutes were kept of a TMG, tria investigator Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 21

54 METHODS group and TSC meetings. The TSC incuded a patient advisor and more detais of their roe in terms of patient and pubic invovement are given in Appendix 1. Further tria foow-up at 3 years After the initia tria resuts were anaysed, funding was obtained to foow up the tria cohort at 3 years. Detais of the methods and resuts for this further foow-up are given in Chapter NIHR Journas Library

55 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 3 Resuts The main resuts from the PACE-UP tria are pubished, 93 and are reproduced here under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided that the origina author and source are credited. Recruitment of participants The CONSORT diagram of participant fow (Figure 3) shows that of the 11,015 peope invited to participate, 6399 did not respond and 548 were excuded as a resut of sef-reported PA guideine achievement; therefore, 1023 out of 10,467 (10%) were randomised. Baseine characteristics of the study popuation (Tabe 4) Participants recruited to the tria were eveny spread across the age groups. Just over one-third of those recruited were men, around two-thirds were married and around one-fifth took part in the tria as a coupe. The majority were in fu- or part-time empoyment, mosty in high-eve manua, administrative or professiona jobs, with a minority in intermediate or routine and manua occupations. About 80% of those recruited were of white ethnicity, around 10% were back/african/caribbean or back British and approximatey 7% were Asian or Asian British. In terms of heath factors, just under 10% were current smokers, around 80% reported their heath as being good or very good, the majority had one or more chronic disease and some sef-reported pain, around 60% reported no current disabiity, around 10% had a high depression score, around 20% had a high anxiety score, and around two-thirds of participants were overweight or obese. Recruitment occurred throughout a four seasons, but was sighty higher in summer and sighty ower in winter. A of these factors were we baanced between the three randomised groups. In terms of objectivey measured baseine PA eves, the nurse-support group had a sighty higher baseineadjusted average daiy step count [7653 steps, standard deviation (SD) 2826 steps] and minutes spent weeky in MVPA in bouts of 10 minutes (105 minutes, SD 116 minutes) than the posta group (7402 steps, SD 2476 steps; 92 minutes, SD 90 minutes) and the contro group (7379 steps, SD 2696 steps; 84 minutes, SD 97 minutes). A higher proportion of the nurse-support group participants were achieving the guideines of 150 minutes per week of MVPA in bouts of 10 minutes (26%, 89/346) than participants in the posta group (20%, 68/339) and those in the contro group (18%, 61/338). The three groups were simiar in terms of average daiy sedentary time, at around 10 hours per day. In terms of sef-reported PA eves the patterns were different, with the contro group reporting the highest number of weeky minutes of MVPA on the IPAQ, not incuding waking; however, the nurse-support group reported higher eves of MVPA if waking was incuded. A sighty higher proportion of participants in the contro group than those in the intervention groups reported being active on the GPPAQ PAI, both excuding and incuding waking. Losses to foow-up Figure 3 shows the osses to foow-up. Of the 1023 peope randomised, 32 (3%) withdrew and eight (1%) were unabe to be contacted at 12 months. In tota, 956 out of 1023 participants (93%) provided at east 1 day of 540 minutes wear time acceerometer data and were incuded in the 12-month primary anayses. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 23

56 RESULTS Enroment Patients aged years from genera practice searches (n = 25,521) Patients eigibe to be invited to participate (n = 21,243) Patients randomy seected to be invited (n = 12,625) Patients invited to participate (n = 11,015; 8886 househods) Excuded (n = 4278) Not eigibe (READ code), n = 4206 In residentia/nursing homes, n = 72 Not randomy seected (n = 8618) Excuded (n = 1610) Not eigibe (GP decision) Patients randomised (n = 1023; 922 househods) Not randomised (n = 9992) No response, n = 6399 Did not wish to take part, n = 2918 Too active, n = 548 Recruited not randomised, n = 127 Aocation Contro group participants (n = 338; 305 househods) Posta intervention participants (n = 339; 307 househods) Nurse intervention participants (n = 346; 310 househods) 3-month foow-up and anaysis Participants anaysed (287 househods) with compete acceerometer data (n = 318) Withdrawn, n = 1 Not abe to be contacted, n = 2 Inadequate acceerometry, n = 17 Participants anaysed (289 househods) with compete acceerometer data (n = 317) Withdrawn, n = 3 Not abe to be contacted, n = 1 Inadequate acceerometry, n = 18 Participants anaysed (286 househods) with compete acceerometer data (n = 319) Withdrawn, n = 8 Not abe to be contacted, n = 3 Inadequate acceerometry, n = month foow-up and anaysis Participants anaysed (292 househods) with compete acceerometer data (n = 323) Withdrawn, n = 3 Not abe to be contacted, n = 3 Inadequate acceerometry, n = 9 Participants anaysed (283 househods) with compete acceerometer data (n = 312) Withdrawn, n = 12 Not abe to be contacted, n = 4 Inadequate acceerometry, n = 11 Participants anaysed (289 househods) with compete acceerometer data (n = 321) Withdrawn, n = 17 Not abe to be contacted, n = 1 Inadequate acceerometry, n = 7 FIGURE 3 The PACE-UP tria CONSORT fow diagram. Adapted from Harris et a. 93 This is an open access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, provided the origina work is propery cited. See: 24 NIHR Journas Library

57 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 4 Baseine characteristics of the 1023 randomised subjects Tria arm Characteristic Contro (N = 338) Posta (N = 339) Nurse (N = 346) Age (years) at randomisation, n (%) (30) 118 (35) 121 (35) (41) 125 (37) 124 (36) (29) 96 (28) 101 (29) Sex (mae), n (%) 115 (34) 124 (37) 128 (37) Marita status (married), n (%) 213 (64) 215 (65) 230 (68) Randomised as a coupe, a n (%) 66 (20) 68 (20) 73 (21) Empoyment status, 79 n (%) In fu- or part-time empoyment 190 (57) 193 (59) 190 (56) Retired 102 (31) 96 (29) 101 (30) Other 39 (12) 39 (12) 50 (15) NS-SEC (current or previous job), 79 n (%) High-eve manageria, administrative, professiona 199 (62) 191 (60) 184 (56) Intermediate occupations 70 (22) 85 (27) 95 (29) Routine and manua occupations 51 (16) 44 (14) 52 (16) Ethnicity, 79 n (%) White 253 (78) 270 (83) 267 (80) Back/African/Caribbean/back British 30 (9) 31 (10) 40 (12) Asian/Asian British 26 (8) 20 (6) 22 (7) Other 15 (5) 4 (1) 6 (2) Current smoker, n (%) 27 (8) 29 (9) 26 (8) Genera heath: 79 very good or good, n (%) 265 (80) 277 (84) 277 (82) Chronic diseases, n (%) None 129 (39) 135 (41) 117 (35) One or two 183 (55) 171 (51) 188 (55) 3 21 (6) 27 (8) 34 (10) Presence of sef-reported pain, 77 n (%) 220 (66) 236 (71) 234 (70) Limiting ong-standing iness, 79 n (%) 76 (23) 73 (22) 74 (22) Townsend disabiity score, 80 n (%) None (0) 190 (57) 196 (59) 210 (62) Sight or some disabiity (1 6) 127 (38) 130 (39) 124 (36) Appreciabe or severe disabiity (7 18) 15 (5) 8 (2) 7 (2) HADS depression score: 75 borderine or high, n (%) 36 (11) 33 (10) 42 (12) HADS anxiety score: 75 borderine or high, n (%) 65 (19) 64 (19) 71 (21) Low sef-efficacy score, 74 n (%) 102 (31) 96 (29) 117 (35) continued Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 25

58 RESULTS TABLE 4 Baseine characteristics of the 1023 randomised subjects (continued) Tria arm Characteristic Contro (N = 338) Posta (N = 339) Nurse (N = 346) Month of baseine measure, n (%) March May 80 (24) 75 (22) 76 (22) June August 105 (31) 106 (31) 110 (32) September November 88 (26) 82 (24) 92 (27) December February 65 (19) 76 (22) 68 (20) Physica characteristics Overweight/obese: BMI of 25 kg/m 2, n (%) 227 (67) 221 (65) 233 (67) Fat mass (kg), mean (SD) 26 (10) 27 (11) 26 (11) Waist circumference (cm), mean (SD) 93 (14) 94 (14) 93 (13) PA data Acceerometry Adjusted baseine step count per day, mean (SD) 7379 (2696) 7402 (2476) 7653 (2826) Tota weeky minutes of MVPA in 10-minute bouts, mean (SD) 84 (97) 92 (90) 105 (116) Average daiy sedentary time (minutes), mean (SD) 613 (68) 614 (71) 619 (78) Average daiy wear time (minutes), mean (SD) 789 (73) 787 (78) 797 (84) 150 minutes of MVPA in 10-minute bouts (yes), n % 61 (18) 68 (20) 89 (26) IPAQ score 72 IPAQ MVPA score: tota weeky minutes of MVPA in 10-minute bouts (N = 909), mean (SD) IPAQ waking score: tota weeky minutes of waking in 10-minute bouts (N = 888), mean (SD) 197 (314) 147 (256) 172 (279) 333 (333) 330 (338) 312 (277) 150 weeky minutes of IPAQ MVPA score (N = 909): yes, n (%) 110 (37) 91 (30) 109 (35) 150 weeky minutes of IPAQ waking score (N = 888): yes, n (%) 193 (65) 190 (66) 208 (69) GPPAQ score, 73 n (%) PAI score (N = 973) Inactive 159 (49) 153 (48) 156 (47) Moderatey inactive 69 (21) 66 (21) 83 (25) Moderatey active 50 (16) 63 (20) 60 (18) Active 44 (14) 36 (11) 34 (10) PAI score, incuding waking (GPPAQ waking score; N = 973) Inactive 129 (40) 134 (42) 133 (40) Moderatey inactive 57 (18) 56 (18) 63 (19) Moderatey active 43 (13) 49 (15) 47 (14) Active 93 (29) 79 (25) 90 (27) SD, standard deviation. a Two and one participants in the posta and nurse groups, respectivey, were randomised and took part in the tria as a coupe, athough their partner was excuded before randomisation because of a ack of wear time. Adapted from Harris et a. 93 This is an open access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, provided the origina work is propery cited. See: 26 NIHR Journas Library

59 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Data competeness for acceerometry Acceerometer wear time was simiar between the groups at baseine, the 3-month foow-up point and the 12-month foow-up point (see Tabes 4 and 5). Over 90% of a groups provided 5 days of 540 minutes wear time at 12 months (see Appendix 2, Tabe 23). Effect of the intervention on acceerometer-assessed physica activity outcomes (Tabe 5) Three-month (interim) outcomes There were significant differences for the change in average daiy step counts from baseine to 3 months between intervention groups and the contro group: additiona step counts (steps per day) were 692 steps (95% CI 363 to 1020 steps; p < 0.001) for the posta group and 1173 steps (95% CI 844 to 1501 steps; p < 0.001) for the nurse-support group, and the difference between the intervention groups was statisticay significant (481 steps, 95% CI 153 to 809 steps; p = 0.004). Findings for the change in time in MVPA eves showed a simiar pattern: additiona MVPA in bouts of 10 minutes (minutes per week) was 43 minutes (95% CI 26 to 60 minutes; p < 0.001) for the posta group and 61 minutes (95% CI 44 to 78 minutes; p < 0.001) for the nurse-support group, and the difference between intervention groups was 18 minutes (95% CI 1 to 35 minutes; p = 0.04). There was no difference between the groups for the change in sedentary time. Summary data for the 3-month PA outcomes are shown in Appendix 2, Tabe 24. TABLE 5 Primary and secondary acceerometry outcome data Comparison between tria arms Posta vs. contro Nurse support vs. contro Nurse support vs. posta Outcome Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue Daiy step count 3 months 692 (363 to 1020) < (844 to 1501) < (153 to 809) months 642 (329 to 955) < (365 to 989) < ( 277 to 349) 0.82 Tota weeky minutes of MVPA in 10-minute bouts 3 months 43 (26 to 60) < (44 to 78) < (1 to 35) months 33 (17 to 49) < (19 to 51) < ( 14 to 17) 0.83 Daiy sedentary time (minutes) 3 months 2 ( 12 to 7) ( 16 to 3) ( 13 to 5) months 1 ( 8 to 10) ( 9 to 9) ( 10 to 8) 0.79 Daiy wear time (minutes) 3 months 2 ( 8 to 12) ( 6 to 14) ( 8 to 12) months 9 ( 1 to 19) ( 0.8 to 19) ( 10 to 10) 0.96 Notes Acceerometry data were avaiabe in the contro, posta and nurse groups for 318, 317 and 319 participants at 3 months, respectivey, and for 323, 312 and 321 at 12 months, respectivey. A modes incude practice, sex, age at randomisation and month of baseine acceerometry as fixed effects and househod as a random effect in a mutieve mode. The xtmixed command in Stata was used, foowed by the post-estimation command pwcompare, to generate the pairwise estimates of effect and their CIs. Adapted from Harris et a. 93 This is an open access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, provided the origina work is propery cited. See: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 27

60 RESULTS Tweve-month (main) outcomes Both intervention groups increased their step counts between baseine and 12 months compared with the contro group: additiona step counts (steps per day) were 642 steps (95% CI 329 to 955 steps; p < 0.001) for the posta group and 677 steps (95% CI 365 to 989 steps; p < 0.001) for the nurse-support group, with no statisticay significant difference between intervention groups (36 steps, 95% CI 277 to 349 steps; p = 0.82). Time spent in MVPA in bouts showed a simiar pattern, that is, both intervention groups increased at 12 months compared with the contro group: additiona MVPA in bouts (minutes per week) was 33 minutes (95% CI 17 to 49 minutes; p < 0.001) for the posta group and 35 minutes (95% CI 19 to 51 minutes; p < 0.001) for the nurse-support group, with no statisticay significant difference between the two intervention groups (2 minutes, 95% CI 14 to 17 minutes; p = 0.83). Again, there was no difference between the groups for the change in sedentary time. Summary data for the 12-month PA outcomes are shown in Appendix 2, Tabe 24. Residuas from the 12-month modes for steps and weeky MVPA in 10-minute bouts were potted, and the distribution of residuas from both modes was normay distributed (see Appendix 2, Figure 18). Effect of the intervention on sef-reported physica activity outcomes at 12 months (Tabe 6) At 12 months, the IPAQ weeky minutes of MVPA (not incuding waking) did not show any effect of the intervention for either intervention group compared with the contro group. However, weeky minutes of waking from the IPAQ at 12 months compared with baseine showed significant increases for both groups compared with contros: 69 minutes (95% CI 19 to 119 minutes) in the posta group and 55 minutes (95% CI 5 to 105 minutes) in the nurse-support group; there was no difference between the intervention groups ( 14 minutes, 95% CI 64 to 37 minutes). This was aso refected in the odds ratio (OR) for achieving 150 minutes of activity in a week, conditiona on the baseine state, which was not significant for IPAQ MVPA for either intervention group, but was for IPAQ waking (posta group vs. contro group: OR 2.1 minutes, 95% CI 1.3 to 3.3 minutes; nurse-support group vs. contro group: OR 1.7 minutes, TABLE 6 Effect estimates for sef-report questionnaires (IPAQ and GPPAQ) at 12 months Comparison between tria arms Posta vs. contro Nurse support vs. contro Nurse support vs. posta Questionnaire outcome Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue IPAQ 72 Change in weeky minutes of activity Vigorous and moderate activity (n = 775) 10 ( 58 to 38) ( 80 to 16) ( 70 to 27) 0.38 Waking (n = 750) 69 (19 to 119) (5 to 105) ( 64 to 37) 0.59 OR for achieving 150 minutes of activity in a week at foow-up, conditiona on baseine state Vigorous and moderate activity (n = 775) 1.0 (0.5 to 2.0) (0.3 to 1.3) (0.3 to 1.3) 0.19 Waking (n = 750) 2.1 (1.3 to 3.3) (1.1 to 2.6) (0.5 to 1.3) 0.41 GPPAQ 73 OR for being active at foow-up, conditiona on baseine state PAI (n = 892) 1.2 (0.7 to 2.1) (0.5 to 1.6) (0.4 to 1.3) 0.32 PAI, incuding waking (n = 892) 1.1 (0.6 to 1.8) (0.5 to 1.5) (0.5 to 1.4) NIHR Journas Library

61 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO % CI 1.1 to 2.6 minutes); there was no difference between the intervention groups (OR 0.8 minutes, 95% CI 0.5 to 1.3 minutes). For the GPPAQ, the OR for being active at the 12-month foow-up, conditiona on the baseine state, did not show a significant effect for either of the intervention groups compared with the contro group, whether or not waking was incuded in the GPPAQ PAI. Summary data for the effect of the intervention on the IPAQ and the GPPAQ are given in Appendix 2, Tabe 25. Effect of the intervention on other heath-reated outcomes (Tabe 7) Fat mass was sighty reduced at 12 months in both intervention groups, but this did not differ significanty from the contro group. There was no change in BMI or waist circumference between baseine and 12 months. The interventions had no significant effects on anxiety, depression, quaity of ife (EQ-5D-5L) or pain scores at either 3 months or 12 months. The exercise sef-efficacy score significanty increased in both intervention groups at 3 months compared with the contro group, and there was a greater effect in the nurse-support group than in the posta group. By 12 months, the sef-efficacy score was significanty higher in the nurse group than in the contro group, and the posta group was intermediate between, but not significanty different from, either of the other groups. Effect of the intervention on adverse events and serious adverse events (Tabe 8) The number of tota AEs did not differ between the groups at either 3 months or 12 months, whether using higher numbers of events sef-reported from the patient questionnaire (fas, fractures, sprains or injuries) or ower numbers of events from primary care records (any AE cardiovascuar, fracture, sprains/ injuries, fas or pain from back or ower imb). There was aso no between-group difference in tria SAEs reported for safety monitoring. Sef-reported fas were ower in the nurse-support group at 12 months (43/318, 14%) than in the posta group (57/310, 18%) or the contro group (71/318, 22%; p = 0.02). Fas reported in primary care records over 12 months were fewer in number, but aso in the same direction, athough the differences were non-significant (p = 0.13). Primary care-recorded cardiovascuar events over 0 12 months were ower in the nurse-support group (2/340, 0.6%) and the posta group (1/331, 0.3%) than in the contro group (8/334, 2.4%; p = 0.04). Subgroup anayses (Figure 4) There was no evidence of effect modification on the change in step count at 12 months for either of the intervention groups versus the contro group for any of the foowing: age, sex, taking part as a coupe, BMI, disabiity, pain, socioeconomic group and exercise sef-efficacy. Sensitivity anayses and imputations (see Appendix 2, Tabe 26) The sensitivity anayses on the primary outcome measure (change in average daiy step count at 12 months), restricted to those with 600 minutes daiy wear time, increased the effect size for both intervention groups versus the contro group, but did not change the interpretation (both intervention groups had a significant effect compared with the contro group, but there was no significant difference between the interventions). Simiary, imputations with both missing-at-random and missing-not-at-random assumptions made some difference to the effect sizes for both interventions compared with the contro group and with each other, but, again, made no difference to the overa interpretation. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 29

62 RESULTS TABLE 7 Effect estimates for other heath-reated outcomes Comparison between tria arms Posta vs. contro Nurse support vs. contro Nurse support vs. posta Heath-reated outcome Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue BMI (kg/m 2 ) 12 months 0.1 ( 0.3 to 0.1) ( 0.2 to 0.1) ( 0.1 to 0.3) 0.42 Fat mass (kg) 12 months 0.4 ( 0.8 to 0.07) ( 0.7 to 0.2) ( 0.3 to 0.6) 0.54 Waist circumference (cm) 12 months 0.04 ( 0.8 to 0.7) ( 0.6 to 0.8) ( 0.6 to 0.8) 0.74 HADS anxiety score 75 3 months 0.3 ( 0.7 to 0.1) ( 0.7 to 0.1) ( 0.4 to 0.4) months 0.2 ( 0.6 to 0.2) ( 0.6 to 0.2) ( 0.4 to 0.4) 1.00 HADS depression score 75 3 months 0.2 ( 0.6 to 0.1) ( 0.5 to 0.1) ( 0.3 to 0.3) months 0.1 ( 0.5 to 0.2) ( 0.4 to 0.3) ( 0.2 to 0.5) 0.51 EQ-5D-5L score 76 3 months ( 0.02 to 0.01) ( 0.03 to 0.01) ( 0.03 to 0.01) months 0.01 ( 0.03 to 0.01) ( 0.03 to 0.01) ( 0.02 to 0.02) 0.87 Exercise sef-efficacy score 74 3 months 1.1 (0.2 to 2.0) (1.4 to 3.2) < (0.3 to 2.1) months 0.6 ( 0.3 to 1.6) (0.3 to 2.2) ( 0.4 to 1.5) 0.22 Sef-reported pain score 3 months 0.05 ( 0.06 to 0.17) ( 0.07 to 0.16) ( 0.12 to 0.11) months 0.05 ( 0.06 to 0.17) ( 0.10 to 0.13) ( 0.15 to 0.08) Notes At baseine, data were avaiabe for a participants for BMI and waist circumference, and for 335, 337 and 346 participants in the contro, posta and nurse-support groups, respectivey, for fat mass. At 12 months, data were avaiabe in the contro, posta and nurse-support groups for 323, 314 and 321 participants, respectivey, for BMI and waist circumference, and for 319, 308 and 320, respectivey, for fat mass. Questionnaire data were avaiabe for varying numbers of participants at baseine, 3 months and 12 months. A modes incude practice, sex, age at randomisation and month of baseine acceerometry as fixed effects and househod as a random effect in a mutieve mode. Adapted from Harris et a. 93 This is an open access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, provided the origina work is propery cited. See: NIHR Journas Library

63 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 8 Adverse events Time frame in each tria arm 0 3 months 0 12 months AEs Contro, n (%) Posta, n (%) Nurse support, Contro, n (%) p-vaue a n (%) Posta, n (%) Nurse support, n (%) p-vaue a AEs reported on the questionnaire (N = 931) (N = 946) Fa, fracture, sprain or injury 59/313 (19) 70/310 (23) 65/308 (21) /318 (36) 99/310 (32) 96/318 (30) 0.34 Fa 25 (8) 24 (8) 24 (8) (22) 57 (18) 43 (14) 0.02 Fracture 3 (1) 3 (1) 7 (2) (5) 10 (3) 11 (3) 0.57 Sprain or injury 49 (16) 54 (17) 47 (15) (21) 68 (22) 63 (20) 0.81 (N = 911) (N = 924) Deterioration in heath probems aready present since the start of the study 33/311 (11) 30/303 (10) 39/297 (13) /313 (22) 67/300 (22) 65/311 (21) 0.91 AEs from primary care records b (N = 1005) (N = 1005) Any AE 29/334 (8.7) 23/331 (7.0) 20/340 (5.9) /334 (25.5) 75/331 (22.7) 77/340 (22.7) 0.62 Cardiovascuar c 2 (0.6) 0 1 (0.3) (2.4) 1 (0.3) 2 (0.6) 0.04 Fracture 4 (1.2) 2 (0.6) 2 (0.6) (3.3) 4 (1.2) 4 (1.2) 0.11 Sprain/injury 2 (0.6) 1 (0.3) 2 (0.6) (2.4) 4 (1.2) 5 (1.5) 0.51 Fa (2.4) 4 (1.2) 2 (0.6) 0.13 Pain (back or ower imb) 23 (6.9) 20 (6.0) 16 (4.7) (19.5) 65 (19.6) 70 (20.6) 0.93 SAE spontaneousy reported d (N = 1023) (N = 1023) SAE 3/338 (0.9) 1/339 (0.3) 3/346 (0.9) /338 (3.0) 5/339 (1.5) 11/346 (3.2) a Chi-squared tests or Fisher exact tests were carried out to assess the statistica significance of the overa differences between the three groups. b A tota of 1005 participants gave permission at randomisation for their primary care records to be accessed and downoaded. c Cardiovascuar events recorded in primary care records incuded a new episode of any of the foowing: myocardia infarction, coronary artery bypass graft, angiopasty, ischaemic heart disease, angina, transient ischaemic attack and stroke. d Information on spontaneousy reported SAEs was coected for the entire cohort (i.e. n = 1023). SAEs were recorded for safety purposes contemporaneousy in the tria, and incuded the foowing: deaths, hospita admission and new-onset disabiity. A of the SAEs reported during the 0- to 12-month tria foow-up were emergency hospita admissions. Adapted from Harris et a. 93 This is an open access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, provided the origina work is propery cited. See: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 31

64 RESULTS (a) Subgroup Treatment effect (95% CI) A participants Gender Femae Mae Age group (years) Coupe No Yes NS-SEC 1, 2 3, 4 5, 6, 7 BMI (kg/m 2 ) < Sef-reported pain None Any Disabiity None Any Sef-efficacy Low High 642 (329 to 955) 552 (169 to 934) 812 (299 to 1325) 639 (202 to 1075) 685 (237 to 1132) 747 (404 to 1091) 141 ( 609 to 891) 606 (201 to 1012) 747 (104 to 1390) 667 ( 179 to 1513) 678 (157 to 1198) 594 (210 to 979) 199 ( 375 to 772) 850 (470 to 1229) 516 (105 to 927) 905 (423 to 1388) 523 ( 58 to 1105) 733 (355 to 1111) Treatment effect (difference between the posta group and the contro group in average daiy step count) at 12 months FIGURE 4 Subgroup anayses. (a) Posta group and contro group; and (b) nurse-support group and contro group. (continued) Summary of the main tria findings Overa, 10% (1023/10,467) of those invited participated in the tria, and we had primary outcome data on 93% (956/1023) of participants at 12 months. Athough the nurse-supported intervention had a greater effect on objective PA outcomes at 3 months, by the main 12-month outcome, both the posta and nurse-supported pedometer interventions significanty increased step counts by around 10%, and time in MVPA in bouts by around one-third compared with the contro group, with no statisticay significant difference between the interventions. There was no significant effect of the interventions on sedentary time or anthropometric measures. In terms of the effects on sef-reported PA eves, the IPAQ MVPA questions did not show any intervention effect, but both the nurse-supported intervention and the posta intervention showed a significant effect on the IPAQ waking question. The GPPAQ score did not show an intervention effect, even when waking was incuded in the score. The interventions had no effect on most other patient-reported outcomes, except that exercise sef-efficacy was increased in both intervention 32 NIHR Journas Library

65 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 (b) Subgroup Treatment effect (95% CI) A participants Gender Femae Mae Age group (years) Coupe No Yes NS-SEC 1, 2 3, 4 5, 6, 7 BMI (kg/m 2 ) < Sef-reported pain None Any Disabiity None Any Sef-efficacy Low High 677 (365 to 989) 554 (171 to 936) 905 (399 to 1411) 776 (344 to 1207) 579 (128 to 1029) 778 (436 to 1121) 188 ( 569 to 946) 545(132 to 959) 793 (163 to 1422) 690 ( 97 to 1478) 806 (271 to 1341) 629 (254 to 1004) 740 (182 to 1297) 658 (278 to 1038) 665 (261 to 1068) 704 (221 to 1187) 705 (157 to 1253) 639 (254 to 1023) Treatment effect (difference between the nurse-support group and the contro group in average daiy step count) at 12 months FIGURE 4 Subgroup anayses. (a) Posta group and contro group; and (b) nurse-support group and contro group. groups at 3 months, and in the nurse-support group at 12 months, compared with the contro group. AEs were not increased by the interventions; some individua AEs were ower in the intervention groups, but this was based on sma numbers of events. No important subgroup effects were demonstrated, and the sensitivity anayses and imputations did not change the interpretation of the tria resuts. The foowing chapters present the resuts reating to other aspects of the tria: the economic evauation (see Chapter 4), generaisabiity and representativeness (see Chapter 5), the process evauation (see Chapter 6), the quaitative evauation (see Chapter 7) and the 3-year tria foow-up (see Chapter 8). Chapter 9 discusses the tria findings in detai. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 33

66

67 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 4 Economic evauation Introduction Evidence on the short- and ong-term cost-effectiveness of pedometer-based interventions coud support the deveopment of poicy and practice encouraging increased PA. Cas for evidence on what works have been made, in both primary and secondary prevention poicy documents, aong with appeas to ink interventions to cear heath outcomes and ensure that resources are used most efficienty. 94,95 To date, ony one pubished estimate of the cost-effectiveness of pedometer programmes for the UK was based on primary evidence. 58 A sma (n = 79), highy seected (80% of women from one GP practice in Gasgow) sampe was used. A maxima pedometer-based waking programme, which incuded two 30-minute consutations (based on the transtheoretica mode of behaviour change), was compared with a waiting ist contro group, which was asked to wait for 12 weeks, after which the group members received a minima waking programme, which incuded a pedometer and two 5-minute sots of brief advice. 45,96 Compared with the 12-week waiting ist contro group, it cost an additiona 92 per person to achieve an additiona eight peope meeting the target of 15,000 steps per day over a 12-week period. Comparing the maxima waking programme with the minima waking programme, it cost a further 591 for one additiona person to achieve the same target. No data were coected on QALYs, and ong-term cost-effectiveness was not modeed. Esewhere, in Austraia, New Zeaand and the Netherands, evidence on the benefits of pedometer-based interventions for primary prevention has been assessed for community-based aduts with ow PA eves Interventions such as pedometer prescriptions and pedometer-based teephone coaching were compared with time-based activity prescriptions or usua practice. Evidence suggests that pedometers may be costeffective in the ong term, but estimates vary widey (from being cost-saving and having fewer disabiityadjusted ife-years in Austraia in the ong term to 11,110 per QALY gained). The generaisabiity of resuts to other contexts has aso not been considered. 100 This chapter examines the short- and ong-term cost-effectiveness, from the NHS perspective, of aternative pedometer-based waking programmes to increase PA eves using PA outcomes for comparison with other PA programmes and QALYs to aid decision-making beyond PA programmes. The interventions are compared against usua practice, in inactive aduts aged years from south London, and are as described in Chapter 2: 1. provision, by post, of pedometers with written instructions 2. pedometers provided with taiored support from a practice nurse. This chapter is structured into two sections: (1) a within-tria anaysis with a time horizon of 1 year, and (2) beyond-tria modeing that takes a ifetime perspective. In each section, the methods and resuts are presented. This is foowed by a discussion of the findings in the context of the strengths and weaknesses of the study, as we as current iterature. Within-tria cost-effectiveness anaysis The popuation, interventions and comparator are identica to Chapter 2. Harris et a. 70 set out the protoco for methods, incuding for the economic evauation. This section covers methods used to measure, vaue and aggregate costs and outcomes, the treatment of missing data and methods of assessing cost-effectiveness. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 35

68 ECONOMIC EVALUATION Methods Identification, measurement, vauation and aggregation of cost To identify NHS resource use, meetings with the tria team estabished who did what, to whom and how often, 101 accounting for events that are ikey to have high unit costs, be frequent or differ between tria arms, and excuding research-focused costs. NHS resources identified for coection incuded: Set-up of service (e.g. design, setting up the intervention in GP practices and staff training, but excuding tria set-up; see Appendix 3, Tabe 27). Deivery of service, incuding, for exampe, pedometers, post/teephone services, handbooks and staff time; a costs fe within months 0 3, except research assistant contacts with participants (4 12 months; see Appendix 3, Tabe 28). Heath service use in primary care (GP and nurse consutations, excuding nurse PA consutations undertaken as part of the tria) and secondary care (hospita admissions, A&E, outpatients), as changes coud occur from treating AEs and changes in ifestye (see Appendix 3, Tabe 29); the heath service data were avaiabe on the 1005 out of 1023 randomised participants who gave written informed consent for their primary care data to be downoaded. Of these, 956 participants (323 in the contro group, 312 in the posta group and 321 in the nurse-support group) aso had 12-month outcome data and, therefore, were incuded in the heath economic anayses reating to heath service use. Resource use was measured using administrative/tria management records, eectronic diaries and interviews of the tria manager and the principa investigator. Participant-eve heath service use (e.g. GP visits, referras) was coected through a one-time downoad of GP records, for those who gave expicit consent for this, at the end of the tria (see Appendix 3, Tabes 27 29). NHS resources were vaued using nationa costs 102,103 (see Appendix 3, Tabes 27 33) to increase generaisabiity. Where nationa unit costs were not avaiabe, oca unit costs from St George s Hospita, London, were used. A costs were expressed in pounds stering, infated to the same base year when appropriate, using the Hospita and Community Heath Service infation index. 103 As the study covers 1 year, costs and outcomes were not discounted. A resources and costs were coected at, or apportioned to, the tria participant eve. The tota cost per participant was the sum of each resource use mutipied by the reevant unit cost over 0 3 and 4 12 months, with costs censored at 12 months. To support a sensitivity anaysis of an aternative wider viewpoint that incuded participants, as participation can be affected by economic barriers, three types of costs borne by participants were coected: participation in the intervention (e.g. time and money spent accessing the intervention for months 0 3; see Appendix 3, Tabe 30), money shown to contribute to the costs of waking and other PA (e.g. membership/event fees, shoes/cothing, food/drink) 104 and money spent as a resut of fas/fractures/sprains/injuries. These data were coected for months 1 3 using participant-competed questionnaires at 3 months and for months using the participant-competed questionnaires at 12 months (see Appendix 1). As the data for months were foow-up data beyond the intervention, costs from months were mutipied by three (to approximate annua costs when added to the cost from months 1 3) and added to the costs from months 1 3 for an annua participant cost. Measurement, vauation and aggregation of outcomes The economic anaysis uses indicators of PA outcomes. The use of cost per additiona step aids comparison of inputs with directy intended and objectivey measured outputs, which the tria was specificay powered to detect, and which therefore reates the economics to the main tria outcome. As objectivey measured weeky minutes of MVPA in bouts of 10 minutes were statisticay significanty different, have important heath impacts, provide a second point of comparison to other studies and ink directy to the onger-term mode, these were aso incuded as an additiona outcome measure for assessing cost-effectiveness. 36 NIHR Journas Library

69 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 To faciitate the comparison of the PACE-UP tria with other heath interventions using the quaity-of-ife measure recommended for evauating heath interventions in Engand, participants aso competed the EQ-5D-5L questionnaire at baseine, 3 and 12 months. The EQ-5D-5L, rather than the EQ-5D-3L (EuroQo-5 Dimensions, three-eve version), was seected as the EQ-5D-3L is known to suffer from ceiing effects and the five-eve version was expected to be more sensitive to differences in heath among heathier peope and, therefore, ess subject to suffer from ceiing effects. This has been subsequenty shown to be the case in Engand, especiay for oder popuations, and, therefore, the EQ-5D-5L has been recommended for use in genera popuation surveys (despite the ceiing effects of dimensions ranging from 58 90%). 105 Utiity weights were assigned at each time point, based on an interim scoring cross-wak function 106 inked to the standard UK-based weights. 107 EuroQo-5 Dimensions (EQ-5D) utiities were converted to QALYs over the tria period using the area under the curve method. 108 Methods of anaysis Missing data Data were investigated for patterns of missingness. 109 Mean imputation was used when the proportion of missing data was 5%. 110,111 Missing EQ-5D-5L data were repaced using an index, rather than domain imputation, as the sampe size was > Mutipe imputation by chained equations was fitted to repace item non-response. In ine with Rubin s rues 113 and other recommendations, 114,115 the point estimate for imputations was derived by averaging estimates of the imputed data based on resuts from five imputations. The point estimate for categorica data was rounded up to the nearest decima point. The imputation mode incuded variabes used in the main mode for the anaysis, whie aso incuding the predictors of missingness. The dependent variabe was incuded in the imputation mode to ensure that the imputed vaues have the same reationship to the dependent variabe as the observed vaues. 116 Incrementa cost-effectiveness anayses Incrementa cost-effectiveness anayses were based on mutipe regression modes to adjust for variations not accounted for by randomisation, and to provide more robust estimates. Generaised inear modes (GLMs) were fitted separatey for costs and QALYs, 117 accounting for the custer effect (identified as househod identifier) via custered standard errors. 118 Modes used for step count and MVPA have been described in Chapter 2. The cost modes used the Poisson distribution and the QALY modes used the binomia 1 famiy, equivaent to beta regression. 119 Athough the generaised inear modes do not account for the correation between costs and QALYs, the efficiency oss (i.e. higher standard errors) wi be minima, as the incusion of the custer effect provides robust standard errors and mitigates the effects of potentia inaccuracies in the famiy distribution used. The choice of distributiona famiy rested on the modified Park test 120 and the comparison of observed and predicted vaues. Covariates incuded the baseine eve (for QALY-based modes), as recommended, 116 practice and variabes found to be correates of PA-reated outcomes that is, demography (age, sex, ethnicity, marita status, education, empoyment, socioeconomic status, cohabitation), heath (number of disease conditions) and other ifestye behaviours (smoking and acoho intake). 121 Reduced modes were generated using Wad tests to examine the joint significance of variabes found to be insignificant in the base mode. Significance eves were set at 5%. To provide more precise estimates of uncertainty, the margins method was used to generate sampe means for tria arms and incrementa point estimates for costs and QALYs. 116,122 A different standard error and cacuated CIs 123 accounted for the custer design. Sensitivity anayses To refect the stochastic uncertainty surrounding the mean incrementa cost-effectiveness, cost-effectiveness panes and cost-effectiveness acceptabiity curves (CEACs) were constructed using 2000 non-parametric sampes from the base-case estimates. Bootstrapping used a new unique identifier for the custers in addition to the origina custer identifier (househod ID). 116 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 37

70 ECONOMIC EVALUATION Deterministic sensitivity anayses performed incuded: a randomised participants (not ony those who provided acceerometry data) varying excusion of costs of heath service use beyond the immediate intervention a method of accounting for AEs ony spontaneousy sef-reported SAEs ony GP data reating to AEs; these were predefined by GP investigators and coaborators (TH, SI, SDW, JI) as possiby being reated to increased waking, and incuded muscuoskeeta events (fas requiring medica attention, fractures, sprains or injuries, pain in back or ower imb) and cardiovascuar events [myocardia infarction, coronary artery bypass graft, angiopasty, stroke, transient ischaemic attack or new-onset ischaemic heart disease (angina)] a perspective of anaysis (i.e. NHS with or without users) variation in the ength of the ife of a pedometer (between 1 and 4 years) scenario combinations (excuding a heath service use costs, and incuding participant costs reated to participation in PA and interventions, except the heath service use cost borne by participants), to ensure that the ong-run mode coud use evidence of worst-case findings. Resuts Tabe 9 summarises the data on costs, EQ-5D-5L utiity scores and quaity of ife (see Appendix 3, Tabes 31 35). At 3 months, the average cost per participant was highest in the nurse group ( 249), foowed by the posta group ( 122) and the contro group ( 107). The mean cost and distribution is affected consideraby by the incusion of heath service use, which resuted in the contro group costing 36 more per participant than the posta group and 12 more than the nurse-support group. QALYs varied marginay, with the gap being the greatest between the contro and posta groups at At 12 months, the average cost per participant was highest in the nurse group ( 603), foowed by the contro group ( 461) and the posta group ( 375). The incusion of heath service use resuted in the contro group costing 86 more per participant than the posta group, but 142 ess than the nurse-support group. QALYs were marginay higher in the posta group (0.843 QALYs) than in the contro group (0.837 QALYs) and the nurse-support group (0.836 QALYs) group. The main resuts (Tabe 10), which are adjusted for baseine differences, show that, at 3 months, the cost of the nurse-support group was statisticay significanty higher than that of the contro group ( 135, 95% CI 99 to 171), whereas this was not the case for the posta group ( 15, 95% CI 15 to 45). Tabe 10 aso shows that there was a statisticay significant increase in daiy steps and minutes of MVPA in 10-minute bouts for the intervention groups compared with the contro group. The ICER per additiona minute of MVPA in 10-minute bouts was 0.35 for the posta group and 2.21 for the nurse-support group, compared with the contro group. However, both intervention groups accrued sighty fewer QALYs than the contro group (posta group: QALYs; nurse-support group QALYs), athough this difference was not statisticay significant. Nevertheess, it contributed to the dominance (ower costs and more QALYs) of the contro group compared with both intervention groups at 3 months. Comparing the two interventions at 3 months (see Tabe 10) shows that the nurse group achieved 481 more steps (95% CI 153 to 809 steps) and 18 more minutes of MVPA (95% CI 1 to 35 minutes) per person than the posta group, at a statisticay significant additiona cost of 120 (95% CI 95 to 146). The estimated cost per additiona step and additiona MVPA minute (in bouts of 10 minutes) was 0.25 and 6.67, respectivey. The nurse-support group had sighty fewer QALYs, athough this was not significanty different ( QALYs, 95% CI to QALYs); however, it contributed to the nurse-support group being dominated (higher costs and fewer QALYs) by the posta group. The main resuts at 12 months (see Tabe 10) were somewhat different from those at 3 months. Athough the mean costs were ower for the posta group ( 91, 95% CI 213 to 33) and higher for the nurse-support group ( 126, 95% CI 37 to 290) than the contro group, neither was statisticay 38 NIHR Journas Library

71 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 9 Average costs and QALYs per participant, by tria arm ( , base case, with missing vaues imputed) Tria arm, mean (SD) Cost and quaity of ife (EQ-5D-5L) Contro Posta Nurse support 0 3 months (N = 318) (N = 317) (N = 319) Costs ( ) Tota cost 107 (254) 122 (107) 249 (215) Set-up cost 0 (0) 45 (0) 105 (0) Deivery of intervention 0 (0) 7 (0) 50 (18) Heath service use 107 (254) 71 (107) 95 (214) Quaity of ife EQ-5D-5L scores at baseine (0.14) (0.12) (0.12) EQ-5D-5L scores at 3 months (0.14) (0.14) (0.14) QALYs, 0 3 months (0.03) (0.03) (0.03) 0 12 months (N = 323) (N = 312) (N = 321) Costs ( ) Tota cost 461 (916) 375(611) 603 (987) Set-up cost 0 (0) 45 (0) 105 (0) Deivery of intervention 0 (0) 10 (0) 52 (18) Heath service use 461 (916) 320 (611) 447 (987) Quaity of ife EQ-5D-5L scores at baseine (0.14) (0.12) (0.13) EQ-5D-5L scores at 3 months (0.14) (0.13) (0.14) EQ-5D-5L scores at 12 months (0.15) (0.13) (0.14) QALYs, 0 12 months (0.13) (0.11) (0.13) TABLE 10 Costs, effects and cost-effectiveness at 3 and 12 months ( ; base case, adjusted for baseine differences) Tria arm, mean (95% CI) Cost, effects or cost-effectiveness Costs and effects over 3 months Contro Posta deivery a Nurse support a Nurse support vs. posta deivery, mean (95% CI) Tota cost per participant ( ) 108 (80 to 136) 123 (111 to135) 244 (221 to 266) Incrementa cost ( ) 15 ( 15 to 45) 135 (99 to 171) 120 (95 to 146) Tota QALYs per participant ( to ) ( to ) ( to ) Incrementa a QALYs ( to ) ( to ) ( to ) continued Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 39

72 ECONOMIC EVALUATION TABLE 10 Costs, effects and cost-effectiveness at 3 and 12 months ( ; base case, adjusted for baseine differences) (continued) Tria arm, mean (95% CI) Cost, effects or cost-effectiveness Contro Posta deivery a Nurse support a Nurse support vs. posta deivery, mean (95% CI) Incrementa daiy steps 692 (363 to 1020) 1172 (844 to 1501) 481 (153 to 809) Incrementa weeky minutes of MVPA in bouts of 10 minutes 43 (26 to 60) 61 (44 to 78) 18 (1 to 35) Costs and effects over 12 months Tota cost per participant ( ) 467 (365 to 569) 376 (307 to 445) 593 (473 to 714) Incrementa cost ( ) 91 ( 215 to 33) 126 ( 37 to 290) 217 (81 to 354) Tota QALYs per participant (0.832 to 0.853) (0.827 to 0.849) (0.824 to 0.847) Incrementa QALYs ( to 0.009) Incrementa daiy steps 642 (329 to 955) ( to 0.007) 677 (365 to 989) ( to 0.011) 36 ( 227 to 349) Incrementa weeky minutes of MVPA in bouts of 10 minutes 33 (17 to 49) 35 (19 to 51) 2 ( 14 to 17) ICER a at 3 months Cost per additiona QALY ( ) Posta deivery dominated by contro Nurse support dominated by contro Nurse support dominated by posta deivery Cost per additiona step count ( ) Cost per additiona minute of MVPA in a bout of 10 minutes ( ) ICER a at 12 months Cost per additiona QALY ( ) Posta deivery is ess costy but has fewer QALYs. 21,162 saved per QALY ost Nurse support dominated by contro Nurse support dominated by post Cost per additiona step count ( ) Posta deivery dominates contro Cost per additiona minute of MVPA in a bout of 10 minutes ( ) Posta deivery dominates contro a For incrementa anayses, the comparisons are posta deivery vs. contro and nurse support vs. contro. significanty different. However, the increase in the cost of moving from posta deivery to nurse-support deivery was statisticay significanty higher ( 217, 95% CI 81 to 354). Athough both interventions were associated with a statisticay significant increase in both step count and weeky minutes of MVPA in 10-minute MVPA bouts compared with the contro group, the difference between intervention groups was not statisticay different at 12 months. The posta group took more steps on average (+ 642 steps) 40 NIHR Journas Library

73 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 and cost ess on average ( 91) than the contro group, and dominated the contro group in terms of the PA outcomes. Compared with the contro group, the nurse-support group cost an additiona 0.19 per step and 3.61 per additiona minute of MVPA in bouts of 10 minutes. None of the sma decrements in QALYs at each incrementa comparison ( QALYs for the posta group vs. the contro group, QALYs for the nurse-support group vs. the contro group and QALYs for the nurse-support group vs. the posta group) was statisticay significanty different. Compared with the contro group, the posta group had fewer QALYs (athough this was not statisticay significanty different) and ower costs (aso not statisticay significanty different). However, the magnitude of the cost-savings is such that they outweigh the forgone QALYs at a threshod of 20,000 per QALY, and woud be considered cost-effective. Using QALYs, the nurse-support group is dominated by both the contro group and the posta group. Comparing the two interventions at 12 months (see Tabe 10) shows that the estimated cost per additiona step and additiona MVPA minute was 6 and 109, respectivey, and that, in terms of QALYs, the nurse-support group was sti dominated by the posta deivery group. The cost-effectiveness panes (Figures 5 and 6; see aso Appendix 3, Figure 19) broady confirm the findings that the posta deivery group had a strong ikeihood of ower costs, but aso fewer QALYs, and that the nurse-support group tended to have fewer QALYs and higher costs than the contro group. However, different eves of uncertainty surround these mean estimates, as refected in the CEACs (Figure 7; see aso Appendix 3, Figure 20). At 20,000 per QALY, the posta deivery group has a 50% chance of being cost-effective compared with the contro group, which fas to 42% at 30,000 per QALY. This is because, as the wiingness-to-pay threshod increases (and, therefore, the higher the vaue that is paced on forgone QALYs), the vaue of QALYs ost begins to outweigh the cost-savings (see Figure 9). This is refected in the CEAC, on which the probabiity moves towards zero. The nurse-support group had ony a 5.5% chance of being cost-effective compared with the contro group at a wiingness-to-pay threshod to gain, or give up, a QALY of 20,000, and this fe to 4.9% when compared with the posta deivery group. 400 Incrementa costs ( ) Incrementa QALYs (EQ-5D) FIGURE 5 Cost-effectiveness pane for the posta deivery group vs. the contro group at 12 months. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 41

74 ECONOMIC EVALUATION 400 Incrementa costs ( ) Incrementa QALYs (EQ-5D) FIGURE 6 Cost-effectiveness pane for the nurse-support group vs. the contro group at 12 months. Probabiity of cost-effectiveness Vaue of threshod ( 000) Intervention group Posta deivery Nurse support FIGURE 7 Cost-effectiveness acceptabiity curve for the posta deivery and nurse-support groups (vs. the contro group) at different wiingness-to-pay-per-qaly threshods. The deterministic sensitivity anayses (see Appendix 3, Tabe 36) show that, with the exceptions of (1) using heath service use incuding ony sef-reported serious adverse effects, (2) excuding a heath service costs, (3) changing perspective (incuding a participant costs) and (4) the worst-case combined scenario, the sensitivity anayses produced resuts that were consistent with the base-case findings. For the exceptions, the posta deivery group was dominated by the usua-care contro group at 12 months. Beyond-tria modeing Systematic reviews have indicated the positive infuence that PA has on primary prevention for a range of conditions, incuding coronary heart disease (CHD), stroke, type 2 diabetes meitus and cancers, and, more recenty, for improving cognition in oder aduts. 129 Carson et a. 5 have separatey shown that increasing amounts of eisure time PA is associated with decreasing heath expenditure. Therefore, by reducing the risk of disease, increased PA can increase future QALYs, as we as ower future costs. The next section provides the methods used to estimate the ong-term cost-effectiveness of the PACE-UP tria. 42 NIHR Journas Library

75 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Methods An existing Markov mode, 130 designed to assess the cost-effectiveness of PA interventions, was used. This mode has evoved from assessing the cost-effectiveness anaysis of exercise referra schemes to brief interventions 129 and beyond. 134 It has been used by NICE to update nationa guidance (PH44 on brief interventions, PH2 on exercise referra schemes guidance) 36,135 and is the basis for the NICE Return on Investment Too 136 used by oca authorities. The mode is driven by evidence of the impact that PA interventions have on the proportion of peope meeting the recommended PA eves, short-term quaity-of-ife gains (associated with meeting the recommended PA eves) and the impact of PA on future rates of CHD, stroke and type 2 diabetes meitus. Figure 8 shows the pathways within the mode. In the origina Markov mode, 129 a cohort of 100, year-od peope were foowed in annua cyces over their ifetime. At the end of the first year of the mode, the cohort woud be either active (doing 150 minutes of MVPA per week) or inactive, and coud have had one of three events (non-fata CHD, non-fata stroke, type 2 diabetes meitus), remain event free (i.e. without CHD, stroke or diabetes meitus) or die, either from cardiovascuar disease (CVD) or from non-cvd-reated causes. Active individuas have a better ife expectancy and quaity of ife, attributabe to ower risks of deveoping CHD, stroke and type 2 diabetes meitus. Peope who become active in the first year (irrespective of the tria arm) accrue a one-off utiity gain associated with achieving the recommended eve of PA. QALYs refect the heath outcomes from a reduced risk of disease. A discount rate of 3.5% per annum is used for costs and QALYs, as recommended by NICE, and the anaysis is conducted from a NHS perspective. Fu detais of the mode are provided in Appendix 3 and esewhere. 130 Cohort of 100,000 heathy but inactive peope (aged 59 years) Run-in period (1 year) Remained inactive Became inactive Remaining ifetime CHD event Event free Stroke Type 2 diabetes meitus Fata Fata Non-fata CHD Non-fata stroke Death resuting from CVD-reated causes Death resuting from non-cvd-reated causes FIGURE 8 Iustration of pathways within the ong-term cost-effectiveness mode. 130 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 43

76 ECONOMIC EVALUATION The mode by Anokye et a. 130 was adapted for use in five ways: 1. The popuation begins with the mean age of the tria popuation (i.e. 59 years rather than 33 years) and was foowed to 88 years (average ife expectancy at 59 years in the UK). 137 This change was aso refected in the age-specific estimates used. 2. The intervention was either pedometer pus nurse support or a posted pedometer. 3. Within-tria costs were used, with a second year of annuitised vaues added for the intervention arms ( 5.03 per person for the posta deivery group and 4.14 per person for the nurse-support group the number of pedometers to the posta deivery group was reativey higher as a resut of more repacements), as the tria anaysis had assumed pedometers asted 2 years. 4. Effectiveness estimates from the PACE-UP tria were used as foows: the probabiity of moving from an inactive to an active state was based on estimated reative risks (RRs) for achieving 150 minutes of MVPA in 10-minute bouts over 7 days at 12 months. RRs of achieving 150 minutes of MVPA in 10-minute bouts at 12 months were estimated from ORs using the formua OR/[(1 P ref ) + (P ref OR)], in which P ref is the proportion of a subjects achieving 150 minutes of MVPA in 10-minute bouts at baseine (i.e. 218/1023 = 0.21). The OR was derived from a ogistic regression mode in which achieving 150 minutes of MVPA in bouts of 10 minutes at 12 months was regressed on baseine minutes of MVPA in bouts of 10 minutes, month of baseine acceerometry, age, sex, genera practice and treatment group, with househod as a custer. 5. The short-term psychoogica benefits associated with achieving 150 minutes of MVPA per week used tria data; incrementa EQ-5D-5L scores (at 12 months) for active peope were regressed, via beta regression, on EQ-5D-5L scores at baseine, ethnicity, education, empoyment status, intervention group, practice and disabiity. Other parameters were informed by estimates from the origina mode. Cost and utiity estimates for disease conditions were originay sourced from iterature reviews of economic evauations conducted for NICE on CVD and diabetes meitus. 5 Estimates for the heath impacts of PA were taken from nationa/ internationa evidence-based guidance on PA and heath, such as the US Physica Activity Guideines Advisory Committee Report, Appendix 3, Tabe 37 detais a parameter vaues. The probabiistic sensitivity anaysis (PSA) addressed the uncertainties around a parameters in the mode (except for baseine mortaity, as the mortaity census data have itte uncertainty). A tota of 10,000 Monte Caro simuations were used to generate stabe estimates. Given that the within-tria sensitivity anaysis showed the decisiona infuence of some assumptions, a deterministic sensitivity anaysis using the ifetime mode expored two aternative, conservative scenarios: Scenario 1 combined excusion of a heath service use costs during the tria period (year 1 of the mode), with no short-term QALY gain associated with achieving the recommended eve of PA. This was considered as a resut of the uncertainty around short-term changes to heath service use, and because previous studies found the excusion of short-term QALY gain associated with being physicay active to the recommended eve to be decisiona. 5,129 Scenario 2 scenario 1 pus user costs reated to participation in PA and the interventions. This combination represented a worst-case scenario in the tria. Athough this perspective is not one adopted by NICE, it represents the most conservative scenario based on this evidence. Resuts Tabe 11 shows that the posta deivery group dominated both usua practice and the nurse-support group, as the ifetime costs were ower and the number of QALYs was greater. The stochastic uncertainty associated with the mean ICER (Appendix 3, Figure 21) indicates that these findings are robust, as there is a 100% ikeihood, at a wiingness-to-pay threshod of 20,000 per QALY, that the posta deivery group is cost-effective compared with the contro group and with the nurse-support group (Figure 9). This is consistent with the net monetary benefit estimates, which show that, athough we can be 95% confident 44 NIHR Journas Library

77 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 11 Costs, effects and cost-effectiveness over a ifetime (cohort of 100,000) Tria arm, mean (95% CI) Costs, effects and cost-effectiveness Contro Posta deivery a Nurse support a posta deivery Nurse support vs. Cost Lifetime tota cost ( M) b 340 (307 to 371) 329 (296 to 361) 351 (318 to 384) Lifetime incrementa cost ( M) 11 ( 12 to 10) 11 (10 to 12) 22 (21 to 23) QALYs Lifetime tota QALYs (miion) 1.07 (0.88 to 1.30) 1.07 (0.88 to 1.30) 1.07 (0.89 to 1.30) Lifetime incrementa QALYs 759 (400 to 1247) 671 (346 to 1071) 108 ( 223 to 10) Lifetime ICER for QALYs ( ) Posta deivery dominates contro 16,368 Posta deivery dominates nurse support Lifetime incrementa net monetary benefit ( 20,000 per QALY) 26 (18 to 36) 2 ( 5 to 11) 24 ( 27 to 21) a For the incrementa anayses, the comparisons are posta deivery vs. contro and nurse support vs. contro. b 46.7M, 37.6M and 59.3M of the tota costs are attributed to the costs of the contro, posta deivery and nurse-supported interventions, respectivey, estimated from the within-tria anaysis. Probabiity of cost-effectiveness Vaue of threshod ( 000) Intervention group Posta deivery Nurse support FIGURE 9 Cost-effectiveness acceptabiity curve showing the probabiity of ifetime cost-effectiveness for the posta deivery group and the nurse-support group (vs. the contro group) at different wiingness-to-pay threshod eves. that the posta deivery group is better than the contro group and the nurse-support group, we cannot be 95% confident that the nurse-support group is better than the contro group. The resuts for scenario 1 of the sensitivity anayses for the 100,000 cohort were as foows: posta deivery group versus contro group posta deivery moved from a dominant position to being a more expensive option (+ 4M) with greater QALY gains (+609 QALYs) and an ICER of 6100 nurse-support group versus contro group the ICER increased from 16,000 to 26,000 (+ 14M, +538 QALYs) nurse-support versus posta deivery group the nurse-support group remained dominated by the posta deivery group (+ 10M, 87 QALYs). Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 45

78 ECONOMIC EVALUATION For scenario 2, the sensitivity anayses for the 100,000 cohort showed the foowing: posta deivery group versus contro group the posta deivery group moved from a dominant position to being more expensive (+ 16M) with greater QALY gains (+609 QALYs) and an ICER of 26,600 nurse-support group versus contro group the ICER increased from 16,000 to 25,400 (+ 13.7M; +538 QALYs) nurse-support group versus posta deivery group the nurse-support group moved from a dominated position to a cost-effective position ( 2M, 87 QALYs). Discussion The within-tria anaysis shows that, at 3 months, compared with the contro group, both interventions cost something to achieve increases in PA. Compared with the contro group, the posta deivery group cost an additiona 0.02 per additiona step gained to an average of steps per day, which was a cheaper buy than the nurse-support group (at 0.12 per additiona step). However, the nurse-support group achieved more steps on average, and the additiona 448 steps were achieved at an incrementa cost of 0.25 per step. Athough this pattern of resuts was repicated for additiona minutes of MVPA (in bouts of 10 minutes), the resuts for QALYs were very different as both intervention groups were dominated by the contro group (i.e. the contro group cost ess and had more QALYs). The main resuts at 12 months were different, eaning more favouraby towards the posta intervention. Compared with the contro group, the posta deivery group achieved statisticay significanty better PA outcomes, and did so at a ower cost. This was much better than the resuts for the nurse-support group, and the insignificant difference in PA outcomes between the nurse-support and posta deivery groups at 12 months impied very high costs per additiona step ( 6.00). The anaysis of the costs per QALY confirmed that the nurse-support group was not a cost-effective aternative. It aso showed that, athough the posta deivery group has fewer QALYs (athough not statisticay different) and ower costs (with costs saved higher than 20,000 per QALY), the posta deivery group coud be considered to be cost-effective. Assuming a vaue of 20,000 per QALY, there was a 50% probabiity that the posta deivery group was cost-effective compared with the contro group, and a 5% probabiity that the nurse-support group was cost-effective compared with the contro group or the posta deivery group at 1 year. The sensitivity anayses did not change the concusions, except in three cases (using sef-reported SAEs, excuding heath service use, but incuding a participant costs), when the posta deivery group was dominated by the usua-care contro group. The ack of evidence on effectiveness in terms of quaity-of-ife outcomes is not necessariy evidence of no effect, as the tria was not powered to detect a change in quaity of ife. The resuts indicate a ot of variation around the change in QALYs (95% CI to QALYs) and we are aware of some ceiing effects at baseine (98% sef-care, 83% usua activities, 79% mobiity, 73% anxiety, 43% pain). Athough this might contribute to raising questions about the reevance of genera quaity-of-ife measures for capturing the quaity-of-ife impacts of pubic heath interventions within the first year, it aso serves to highight the importance of capturing the QALY impact of pubic heath interventions on disease avoidance in onger-term economic modes. Cost-per-QALY resuts from short-term pubic heath trias have the potentia to misead decision-makers on the efficiency of investments in the context of changes that ead to onger-term reductions in the risk of disease. A ifetime cost-effectiveness mode characterised the ong-term impact of PA interventions on CHD, stroke and type 2 diabetes meitus. 129 This showed that the posta deivery group woud dominate both the contro group and the nurse-support groups, as quaity-of-ife gains (759 QALYs, 95%CI 400 to 1247 QALYs) add to increased cost-savings ( 11M, 95% CI 12M to 10M), resuting in an incrementa net monetary benefit of 26M (95% CI 18M to 36M) for a 100,000 cohort. There was a 100% ikeihood that the posta intervention was cost-effective compared with the contro group and the nurse-support 46 NIHR Journas Library

79 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 group. The conservativey framed deterministic scenario anayses showed that excuding both the short-term reduction in heath service use and the utiity gain seen in the tria woud not ater the main concusion that the posta deivery group woud be an extremey cost-effective intervention (ICER of 6100 per QALY). Even taking the unusua step of incuding participant costs did not raise the ICER beyond a threshod vaue of 30,000 per QALY. Strengths and weaknesses To our knowedge, this is the first pubished study of short- and ong-term costs and effects of a pedometer intervention. Its strengths are the use of detaied individua-eve cost and effectiveness data from a we-designed popuation-based RCT, which had neary compete (93.4%) foow-up data to 1 year, to estimate the cost per QALY at 1 year, and aso to input into a ifetime mode of cost-effectiveness. It aso incuded both provider and user perspectives in costing service provision and participation, which aowed the impact of perspective to be investigated, and which provides a basis for further investigation of the roe of cost in its association with participation in PA. The estimates of uncertainty extended commony used techniques to account for increased precision in the context of custered data. The sensitivity anaysis pushed both the short- and ong-term anayses to very conservative outcomes given the tria data and, therefore, provides a good indication of the robustness of findings based on the current evidence. The weakness of the within-tria cost-effectiveness study connects to the cost of heath service use over the period of the tria; no information was avaiabe on the severity or procedures used for hospita admissions, or cause of admission to the A&E department. We reied on the principa investigator s best guess or nearest appropriate code (whie bind to the treatment group) and on averaging across eective/ non-eective admissions and we therefore expored aternative assumptions in our sensitivity anayses. There was considerabe variation in costs in each tria arm, and the tria was not powered to detect a difference. Data were aso not coected on costs to participants for months 4 9, and the ast 3 months were mutipied to represent the missing data. These may have over- or underestimated participant costs and, if significanty underestimated, coud be decisionay important. With respect to the ong-term modeing, the mode assumes that peope woud revert to PA patterns observed in ong-term cohort studies. This estimate coud be improved with onger-term tria data. A the chaenges set out in previous work 130 are aso reevant here (e.g. some diseases, such as cancer and AEs, are not accounted for, which coud ead to either over- or under-estimation of cost-effectiveness). The study feeds into an area sparse of primary data 139,140 popuated ony by sma studies. 95,96 Leung et a. 97 showed a 95% ikeihood that pedometers woud be a cost-effective addition to green prescriptions (in New Zeaand) at 12 months, which is much higher than the 50% ikeihood that we found. Our study aso provides ong-term estimates based on the popuation-eve primary data for comparison with the arger body of cost-effectiveness estimates 134 from decision modes. 97,98,141,142 Some have identified cost-savings and an improved quaity of ife at a popuation eve from pedometers in the ong term. 98,137 Others 97,138 have indicated high probabiities that pedometers wi be cost-effective in the ong term, with Brennan et a. 142 indicating that, even with ong-term support at 25 per year (for monitoring and support), ICERs fe we beow 10,000 per QALY gained. Our study provides further support to indicate that pedometer-based programmes are a cost-effective method of improving heath. Concusion A range of sensitivity anayses of both short- and ong-term cost-effectiveness confirmed the view that posta deivery of pedometer interventions to peope aged years through primary care has a high chance of being cost-effective in the ong term and has a 50% ikeihood of being cost-effective, through resource-savings from changes in heath service use, within 1 year. Further research is needed to ascertain the eve of maintenance of PA beyond 1 year and the impact of PA on quaity of ife and genera heath service use in both the short term and ong term. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 47

80

81 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 5 Generaisabiity Introduction Athough the numbers of aduts achieving the recommended PA eves are generay ow, 9 there are marked differences between groups, with ower PA eves in women, oder peope and those from socioeconomicay deprived areas and of Asian ethnicity. 9,143 Waking interventions aiming to increase PA eves ideay need to try to ensure that these groups are we represented. When participation rates are ow, there may be systematic differences between those who participate and those who do not, but to whom the intervention coud reasonaby be appied. Faiure to incude certain groups of peope for whom PA eves are ower may ead to the impementation of interventions that are ikey to increase heath inequaities. When there are differences between participants and non-participants, exporing the reasons for non-participation using a quaitative approach can be instructive. The popuation-based samping frame used in this study provided an opportunity to assess whether or not there were differences in terms of age, sex, ethnicity and area-eve deprivation between genera practice patients who repied to the invitation etter compared with those who did not. We aso compared the heath, ifestye, education and socia factors of those who agreed to participate in the tria with those who agreed to compete a questionnaire, but who did not wish to participate (see Appendix 4 for the non-participant questionnaire). These findings are now pubished. 147 Quaitative interviews with a sampe of this atter group aowed us to investigate the reasons for non-participation in the tria, and these findings are aso pubished. 148 Methods Data coection for quantitative comparisons The sex, age and IMD score of a those invited were coected from genera practice records. The IMD score is an anonymised measure of deprivation based on postcode. 71 To avoid the possibiity of individuas being identified, aggregated practice-recorded ethnicity was exported from the practice in 10-year age bands for a batches where everyone was maied, ess excusions (n = 10,155). Those not wanting to participate in the tria were asked in their invitation etter if they were wiing to compete a shortened version of the tria baseine questionnaire, incuding questions on demographics, heath, PA eves and a question on reasons for not participating. The foowing categories were offered, based on previous research, with space to add other reasons for tria non-participation: (1) I do not have time, (2) I cannot increase my PA, (3) I am not interested in increasing my PA, (4) I am aready very physicay active, (5) I am not interested in research and (6) I do not want to be put in a group by chance. Comparison groups Individuas whose invitation etters were returned to sender were excuded from the anayses before cacuating response rates. Responders are defined as those who repied to the invitation etter, regardess of whether or not they wanted to take part. Individuas coud respond by post, e-mai or teephone. Participants are those who competed the baseine assessment, athough not a were randomised as some provided inadequate acceerometry data. Non-participants are those who competed a questionnaire but did not wish to participate in the tria (Figure 10). As the PACE-UP tria targeted inactive aduts, participants who attended a baseine appointment were seected on the basis of their ow PA eves. Non-participants were not seected in this way. In order to minimise seection bias, the quantitative anaysis of participants and non-participants was therefore restricted to those categorised as not active according to a sef-reported primary care PA questionnaire, the GPPAQ, 73 which was the ony PA measure avaiabe for both groups. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 49

82 GENERALISABILITY Seected for study (n = 12,625) Sent invitation etter (n = 11,015) Invitees Age and sex matched with GP records (n = 10,927) Excuded (GP decision), n = 1421 Returned to sender, n = 189 Unabe to match for age and sex (n = 88) Responders Sent a repy to invitation (n = 4572; 42%) Non-responders No repy to invitation (n = 6355; 58%) Anaysis of non-responders Wanted to take part in tria (n = 1860; 17%) Did not want to take part in tria (n = 2712; 25%) Not randomised, n = 710 Excuded as too active, n = 548 No baseine assessment, n = 162 Did not compete questionnaire (n = 1573) Participants Competed a questionnaire (n = 1150; 11%) Non-participants Competed a questionnaire (n = 1139; 10%) GPPAQ active, n = 118 GPPAQ missing, n = 108 GPPAQ active, n = 388 GPPAQ missing, n = 36 Not active participants (n = 924) Not active non-participants (n = 715) Anaysis of participants and non-participants FIGURE 10 Fow chart to show the recruitment process in the PACE-UP tria. A percentages are out of a of those whose age and sex were matched with GP records (n = 10,927). Statistica anaysis Age- and sex-standardised rates were used to compare IMD quinties for responders. Simiary, sex-standardised rates were used to compare age groups and age-standardised rates were used to compare sexes. The fu popuation of invitees was used as a standard popuation throughout. No further anaysis on non-responders was possibe because they did not provide any questionnaire data on ethnicity or other factors. 50 NIHR Journas Library

83 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Practice ethnicity data were avaiabe in 10-year age bands for 10,155 invitees, which was effectivey a random sampe of the 11,015 peope invited to participate. The proportion of patients beonging to each ethnicity category within each age band and within each practice was cacuated, and the number of invitees in each ethnicity in each practice and age band was estimated. Overa, 1903 invitees had ethnicity recorded as unknown. These are assumed to be missing at random in the main resuts, but sensitivity anayses were performed assuming that these were a white peope or a non-white peope. Age-standardised participation rates for not-active participants and non-participants competing questionnaires were cacuated, assuming that invitees gave the same ethnicity on the questionnaire as was recorded in their practice records. Participation rates by age, sex and IMD score were cacuated for not-active participants versus not-active non-participants competing questionnaires, as in the anaysis of responders. Not-active participants and non-participants competing the questionnaire were compared for additiona demographic and socia characteristics, and heath and ifestye factors, using ogistic regression. A data came from the questionnaires. Modes were adjusted for custering by practice and househod, by incuding fixed effects for practice and using robust standard errors for househod. Methods for the interview study of non-participants This is fuy described esewhere. 148 Non-participants competing the questionnaire were asked if they coud be contacted to discuss their reasons in more detai. A purposive sampe of those wiing to be contacted was seected to provide men and women of varying ages, ethnicities and empoyment status from the initia six participating practices. To maximise participation, we used focused teephone interviews; permission was gained for interviews to be audio-recorded. The topic guide was deveoped from the iterature, from the previous PACE-LIFT tria quaitative evauation 21,149 and discussions between the authors, and is pubished 148 and provided in Appendix 4. Approximatey 30 interviews were panned, with recruitment continuing unti no new themes were identified and a demographicay baanced sampe had been achieved. Open questions were asked about what infuenced their decision not to participate and their opinions on the tria information received. Responses given on the competed questionnaires were used as a starting point to further expore their reasons. They were asked genera questions about the perception of the tria design and were invited to make any fina comments. Methods for anaysing interviews Audio-recordings were transcribed verbatim and checked for accuracy. After 10 interviews, researchers (Rebecca Homes, TH and CV) read the transcripts and discussed the interviews. The interview technique was then modified sighty to ensure that interviewees understood the tria randomisation process, as severa participants had appeared not to understand the question about whether or not being put in a group by chance had infuenced their decision not to participate. On competion of the interviewing, the transcripts were read and re-read for famiiarisation by the researchers, who assigned codes (RN and TH), before a thematic framework was produced. 150 Coding discrepancies between researchers were resoved by discussion. The framework produced was informed both by a priori issues, mosty reated to tria design, and by emerging themes. Themes were refined further by discussion between authors (Rebecca Homes, TH, RN and CV), and broader categories, encompassing severa subthemes, were generated. Reasons for decining given by a non-participants were aso compared with those given at interview, to put our findings in a wider context and assess the generaisabiity to a of those activey decining. Resuts Resuts from quantitative comparisons Of the 12,625 individuas seected for screening (see Figure 10), 1421 (11.3%) were excuded by practice staff and 189 (1.5%) had invitation etters that were returned, as they had moved away; both of these groups were cassified as not invited. In the 44 househods where one person refused the invitation and the other did not respond, it was impossibe to match the response to individua invitees within the househod, so age and sex are unknown. These 88 peope have been excuded from further anayses. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 51

84 GENERALISABILITY Of the remaining 10,927, 4572 (42%) responded to the invitation etter, mainy by post, and 1150 (11%) competed the baseine assessments. Of a invitees, 5229 (48%) were aged years. Athough a quinties of deprivation were represented, ony 7% were in the most deprived quintie. Response rates were higher in oder peope, women and those iving in ess deprived areas (Tabe 12). As individua ethnicity was avaiabe ony for the participants and non-participants who competed a questionnaire, it was not possibe to estimate response rates by ethnicity for a responders. Athough the GPPAQ was not used to assess PA eves for tria incusion, it was the ony PA measure avaiabe for both participants and non-participants. A tota of 118 participants and 388 non-participants were cassified as active by the GPPAQ, and 134 peope did not compete the GPPAQ. These peope were excuded from further anaysis, eaving 924 participants and 715 non-participants. Simiar to the response rates, participation rates were higher in oder peope, women and those iving in ess deprived areas. (Tabe 13). Ethnicity was extracted from the practice for 10,155 invitees. Of these, 5991 were recorded as white (59%), 893 (9%) were recorded as Asian or British Asian and 915 (9%) were recorded as back Caribbean, back British or back African. A tota of 1903 (18.7%) were recorded as unknown. The percentage of peope with unknown ethnicity varied by practice from 3% to 48%. Tabe 13 shows the estimated number in each ethnicity category for a 10,927 invitees, assuming that those for whom ethnicity was recorded as unknown and those for whom we were not abe to coect ethnicity have the same ethnicity distribution as the group with known ethnicity. TABLE 12 Responders to the invitation etter by age, sex and IMD quintie Responders to the invitation (N = 4572) Characteristic A invitees (N = 10,927), n (%) n Standardised percentage response a (95% CI) Ratio of response rates (95% CI) Age (years) Sex (47.8) (32.1 to 34.7) 0.57 (0.54 to 0.60) (30.8) (44.5 to 47.9) 0.79 (0.76 to 0.84) (21.3) (55.8 to 59.8) 1.00 Femae 5604 (51.3) (45.4 to 48.0) 1.00 Mae 5323 (48.7) (35.5 to 38.1) 0.80 (0.76 to 0.84) IMD nationa quintie b 1 (most deprived) 712 (6.8) (26.2 to 32.8) 0.55 (0.50 to 0.61) (26.4) (34.4 to 37.9) 0.67 (0.63 to 0.72) (28.2) (39.8 to 43.2) 0.77 (0.73 to 0.82) (22.2) (43.6 to 47.5) 0.85 (0.80 to 0.90) 5 (east deprived) 1711 (16.3) (51.4 to 56.0) 1.00 a Age percentages were standardised for sex, sex percentages were standardised for age, and IMD percentages were standardised for age and sex. Percentages are of a those invited. b A tota of 448 peope are missing IMD quintie, primariy because of certain postcode areas not being incuded in the ook-up tabe. 52 NIHR Journas Library

85 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 13 Competion of the baseine assessment and questionnaires in participants and non-participants who are not active on the GPPAQ, by age, sex, IMD quintie and ethnicity Participants (N = 9241) a Non-participants (N = 7152) b Characteristic A invitees (N = 10,927), n n Standardised competion rate a (95% CI) Ratio of competion rates (95% CI) n Standardised competion rate c (95% CI) Ratio of competion rates (95% CI) Age (years) (5.7 to 7.1) 0.60 (0.51 to 0.71) (4.0 to 5.1) 0.41 (0.34 to 0.49) (9.1 to 11.1) 0.94 (0.81 to 1.10) (5.5 to 7.1) 0.56 (0.47 to 0.67) Sex (9.4 to 11.9) Femae (9.8 to 11.4) (10.0 to 12.6) (6.5 to 7.9) Mae (5.6 to 6.9) 0.59 (0.52 to 0.67) (5.2 to 6.5) 0.82 (0.71 to 0.94) IMD nationa quintie d 1 (most deprived) (3.8 to 7.2) 0.52 (0.39 to 0.70) (3.0 to 6.0) 0.51 (0.37 to 0.70) (5.7 to 7.6) 0.63 (0.52 to 0.78) (3.8 to 5.4) 0.52 (0.41 to 0.66) (8.6 to 10.7) 0.92 (0.77 to 1.10) (6.2 to 8.0) 0.80 (0.65 to 0.98) (7.7 to 10.0) 0.84 (0.69 to 1.02) (6.3 to 8.4) 0.83 (0.67 to 1.03) 5 (east deprived) (9.1 to 11.9) (7.5 to 10.2) 1.00 Ethnicity White 81,295 e (8.1 to 9.3) (7.3 to 8.4) 1.00 Asian 11,315 e (4.1 to 6.7) 0.62 (0.50 to 0.76) (1.5 to 3.3) 0.31 (0.24 to 0.38) Back 10,845 e (6.7 to 10.2) 0.97 (0.79 to 1.20) (1.1 to 2.8) 0.24 (0.19 to 0.31) Other 5835 e (2.2 to 5.4) 0.44 (0.33 to 0.59) (2.2 to 5.6) 0.59 (0.36 to 0.68) a Participants competed the baseine heath and ifestye questionnaire and the baseine assessment. b Non-participants competed the non-participant questionnaire. c Age percentages were standardised for sex, sex and ethnicity percentages were standardised for age, and IMD percentages were standardised for age and sex. Percentages are of a those invited. d A tota of 448 peope are missing IMD quintie data, primariy because of certain postcode areas not being incuded in the ook-up tabe. e Number of invitees estimated from practice summary data. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 53

86 GENERALISABILITY Of the white invitees, 709 (8.7%) agreed to participate in the tria and were not active, and a further 638 (7.9%) competed a non-participant questionnaire and were not active. Both Asian and back invitees had very ow non-participant questionnaire competion (2.4% and 1.9%), but back invitees were as wiing to participate as white invitees (8.5% vs. 8.7%), whereas ony 5.4% of Asian invitees participated. The sensitivity anayses assuming that a ethnicities recorded as unknown were white or non-white peope showed simiar resuts, and the same patterns were aso seen in practices with neary compete ethnicity coding. Those providing questionnaire data were more ikey to be working part-time, married or iving with a partner, and ess ikey to have finished their education aged 16 years (Tabe 14). Participation was associated with recent primary care contact and with some degree of heath probems (genera heath, ong-standing iness and comorbidities), athough those more severey affected were ess ikey to participate (see Tabe 14). This is consistent with the EQ-5D-5L (heath-reated quaity-of-ife) domains, whereby participants were more ikey to have probems with pain and mobiity, but ess ikey to have probems with sef-care, which is ikey to indicate greater disabiity. Forty-five per cent of the sampe gave insufficient time (n = 327) or aready being physicay active (n = 325) as reasons for non-participation, even though those cassified on the GPPAQ as active were excuded from this anaysis. Less commony, 152 (21%) peope answered that they coud not or were not interested in (n = 122, 17%) increasing their PA. Randomisation was cited as a reason for non-participation ony by 88 respondents (12%). TABLE 14 Participants and non-participants who competed questionnaires and were not active on the GPPAQ: demographics and heath and ifestye factors Variabes Participants with baseine information (N = 924), a n (%) Non-participants who competed a questionnaire (N = 715), a n (%) OR for participation adjusted for custering (95% CI) b OR for participation adjusted for custering, age and sex (95% CI) Demographic factors Invited as a coupe 393 (42.3) 314 (43.9) 0.98 (0.79 to 1.21) 0.99 (0.79 to 1.23) Married/iving together as a coupe 595 (65.8) 439 (62.5) 1.20 (0.96 to 1.49) 1.25 (1.01 to 1.56)* Age (years) finished fu-time education (26.4) 246 (35.6) 0.64 (0.49 to 0.83) 0.67 (0.51 to 0.87) 17 or (22.6) 122 (16.2) 1.39 (1.10 to 1.76) 1.23 (0.93 to 1.64) (48.3) 334 (48.3) 1.0** 1.0** Empoyment status Fu-time 334 (37.1) 248 (35.4) 1.0*** 1.0** Part-time 175 (19.4) 83 (11.8) 1.60 (1.17 to 2.19) 1.57 (1.13 to 2.18) Retired 274 (30.4) 269 (38.4) 0.77 (0.60 to 0.99) 0.87 (0.63 to 1.21) Other 118 (13.1) 101 (14.4) 0.87 (0.64 to 1.19) 0.85 (0.62 to 1.17) Residentia status Home owner 734 (82.7) 587 (84.2) 0.92 (0.69 to 1.23) 0.91 (0.68 to 1.21) Heath and ifestye factors Contact with GP or nurse in the ast 3 months 591 (65.4) 409 (59.3) 1.31 (1.61 to 1.06)* 1.34 (1.09 to 1.65)** Current smoker 74 (8.4) 62 (9.0) 0.90 (0.63 to 1.29) 0.87 (0.60 to 1.24) 54 NIHR Journas Library

87 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 14 Participants and non-participants who competed questionnaires and were not active on the GPPAQ: demographics and heath and ifestye factors (continued) Variabes Participants with baseine information (N = 924), a n (%) Non-participants who competed a questionnaire (N = 715), a n (%) OR for participation adjusted for custering (95% CI) b OR for participation adjusted for custering, age and sex (95% CI) Genera heath eve Very good/good 727 (81.0) 579 (84.0) 1.0* 1.0* Fair 154 (17.2) 88 (12.8) 1.34 (1.01 to 1.79) 1.40 (1.05 to 1.86) Poor/very poor 16 (1.8) 22 (3.1) 0.54 (0.28 to 1.04) 0.56 (0.29 to 1.09) Limiting ong-standing iness Yes, a ot 24 (2.7) 46 (6.7) 0.40 (0.24 to 0.66) 0.41 (0.24 to 0.70) Yes, a itte 194 (21.7) 113 (16.4) 1.35 (1.04 to 1.77) 1.40 (1.07 to 1.84) No 678 (75.7) 528 (76.9) 1.0*** 1.0*** One or more comorbidities One or more different medications taken per day 568 (58.6) 401 (41.4) 1.23 (1.01 to 1.51)* 1.29(1.05 to 1.59)* 517 (57.6) 384 (55.5) 1.07 (0.87 to 1.31) 1.17 (0.95 to 1.46) EQ-5D measurements Mobiity some probems Sef-care some probems Usua activities some probems Pain/discomfort some probems Anxiety/depression some probems 202 (22.4) 122 (17.4) 1.36 (1.05 to 1.76)* 1.44 (1.10 to 1.87)** 23 (2.6) 31 (4.4) 0.53 (0.30 to 0.93)* 0.56 (0.32 to 0.99)* 163 (18.3) 121 (17.2) 1.06 (0.81 to 1.38) 1.09 (0.83 to 1.43) 522 (58.0) 326 (46.4) 1.61 (1.31 to 1.97)*** 1.62 (1.32 to 2.00)*** 247 (27.8) 169 (24.0) 1.20 (0.96 to 1.52) 1.19 (0.94 to 1.50) Heath factors reating to exercise Baance probems 106 (11.7) 64 (9.3) 1.26 (0.91 to 1.76) 1.27 (0.90 to 1.78) One or more fas in the past year Brisk/fast waking pace 157 (17.5) 123 (18.0) 0.97 (0.74 to 1.26) 0.98 (0.75 to 1.27) 256 (27.9) 342 (48.2) 0.42 (0.34 to 0.51)*** 0.39 (0.32 to 0.49)*** Someone to wak with Sometimes/often/ aways 791 (87.2) 600 (84.2) 1.25 (0.93 to 1.69) 1.20 (0.88 to 1.63) *p < 0.05, **p < 0.01, ***p < 0.001, from the Wad test p-vaue for the incusion of the variabe in the ogistic mode, used to assess the significance of incusion of categorica variabes with more than two categories. a Tota number in each group. Some questions have missing data. b ORs are from modes with fixed effects for practice and robust standard errors for custering by househod. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 55

88 GENERALISABILITY Resuts from quaitative comparisons Fifty-five tria non-participants were teephoned from March to Juy 2013: 21 coud not be contacted and four decined to be interviewed. Thirty tria non-participants representing the six initia participating practices were interviewed. Data saturation was achieved prior to competing 30 interviews, but we continued to 30 to ensure a more ethnicay diverse sampe. The demographic detais of interviewees and their main reasons for non-participation are pubished. 141 Those interviewed were not seected on the basis of inactivity, and a sighty higher percentage (67%; 20/30) reported being too active as a reason for non-participation compared with those who were incuded in the main quantitative anaysis. Most interviewees gave one primary reason for decining participation, consistent across sex, ethnicity and age groups. The majority (n = 18) said that they were too active either because they fet that their activity exceeded the tria s target eves or because others woud benefit more. Less frequenty cited main reasons incuded existing medica probems (n = 4), trave from home (n = 3), work/other commitments (n = 3), concerns about potentia equipment probems (n = 1) and reuctance to be randomised (n = 1). To further understand the reasons for non-participation, we categorised the emerging themes into three domains: interna, externa and tria reated. Short quotations iustrating a of these reasons are shown in Tabe 15 and more detaied quotations are given in our pubished paper. 148 TABLE 15 Summary of categories and themes: barriers to participating Category Subcategory Theme Quotations with non-participant (NP) number Interna Aready active Persona activity I tend to wak quite a ot anyway, so I didn't think a pedometer woud probaby be ikey to increase my waking at a reay. NP12 Work activity I actuay work as a postman, so I do a he of a ot of waking... and that was basicay the reason that I didn t think I d need to actuay join the programme. NP06 Medica probems Stroke I had the stroke in 94. So that imited my waking. NP07 Pain If I wak for more than haf an hour at a time, I get incrediby stiff and painfu. NP16 Heart condition I m aways at the hospita seeing a cardioogist. NP18 No wish to increase activity Mutipe medica probems Not interested/does not ike PA I don t need anything ese going on to do with heath... I certainy woud have thought... that they woud have thought, oh, she woud not want to do this because she s got ots of other probems. NP18 And I do not reay ike running... and I certainy wi not join a gym. I hate exercising. NP02 Aready doing enough No, I think I do enough. I m fine with what I do. NP17 Not interested in waking More interesting than waking Team sport Cycing s nice, swimming... any form of recreation thing, ike ice skating or horse riding or bicyce riding, anything ike that... Waking s quite boring. NP02 We, it woud be hard for you to organise team sports I shoud think woudn t it? I mean I used to pay badminton quite a ot which I enjoyed. NP19 56 NIHR Journas Library

89 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 15 Summary of categories and themes: barriers to participating (continued) Category Subcategory Theme Quotations with non-participant (NP) number Running If anybody s doing research into peope who have had heart attacks and then trying to get back into running, that I d be extremey interested in. NP24 Not the right person For younger peope You get to a stage in your ife and you think, that s it... I m reaxing now. I exercise my mind instead. NP02 For oney peope For overweight peope These sort of things peope take them up if they are oney and I m not oney. NP18 Uness you were a reay fat person, which I m not. NP18 Atruism... an opportunity for someone ese, you know, that it may be more usefu to. NP13 Externa Work commitments It s bad enough trying to get... a day off for a norma appointment. NP02 I just did not think I d have time...because I know how important waking is, and I ove waking, and if I have an hour or two free, I woud prefer to wak than tak to the nurse. NP21 Trave difficuties If I had time, I d ove to be part of your research and go to the surgery and a the rest of it, but I think, actuay... the awkwardness of the journey... NP22 Other commitments Trave from home I m going away so much, I coud not reay tie mysef down to anything ike that. NP01 Caring for famiy member Chores/ ife I m a carer for my father. I think most of it is just being there. NP04 I ve got grandchidren. I ve got a husband. I ike to do my gardening. I ve got a four bedroom house to keep cean. I fee my oad is more than enough to keep me going. NP08 Tria reated Advice from others Length of programme I did mention it to my daughter actuay and she said that sounds crazy! She said it s not for me, so I did not go any further. NP07 It does sound a bit on the engthy side does not it reay... some peope coud be put off by that. NP10 Tria materia Too ong... there was a ot to read. Buet points are good. Just make it simpe. NP19 Aimed at oder peope I just remember thinking, actuay, I do not think I m in that age group yet. It kind of seemed to be geared to peope who reay were in their 70s and over. NP09 continued Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 57

90 GENERALISABILITY TABLE 15 Summary of categories and themes: barriers to participating (continued) Category Subcategory Theme Quotations with non-participant (NP) number Equipment probems Pedometer/ acceerometer We I mean I have actuay used a pedometer but I woud not sort of particuary want to do it for a week. NP09 Randomisation Did not ike concept I think if you re doing research then you shoud be abe to choose... within reason... what cub you re wiing to join reay. NP13 Venue Did not want to be in nurse-support group Did not want to be in the contro group Fitness-reated venue better Does not ike the GP surgery... I coud probaby commit to the other two groups, but possiby not to the nurse support. NP09 We... I coud not see the point of being in a group that did nothing. NP04 If you re going to do a fitness programme, you shoud do it in a fitness venue. NP04 I have to go there when I m not we. I certainy am not going to go to the surgery when I m we. NP18 Waking environment Boring Waking s quite boring. Uness you re waking somewhere on an outing somewhere, you know, in the country or something, seaside. You shoud have more trips. NP02 Wrong season As the weather gets better, then I might go for a wak in the evening... it was reay due to the seasons as we. NP28 Preferred group I think you get more encouragement if you are in a group. NP05 Tria design... that is not something I wanted to be part of I think I d have found it incrediby boring. NP18 Adapted from Normanse et a. 148 This artice is distributed under the terms of the Creative Commons Attribution 4.0 Internationa License ( which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the origina author(s) and the source, provide a ink to the Creative Commons icense, and indicate if changes were made. The Creative Commons Pubic Domain Dedication waiver ( appies to the data made avaiabe in this artice, uness otherwise stated. Interna reasons for non-participation incuded aready being active, medica probems (pain, heart conditions, stroke and mutimorbidity), no wish to increase activity, no interest in waking, feeing incorrecty targeted and atruistic reasons. The dominant reason in this category was a beief in aready being sufficienty active. When expored in more depth, it seemed that, on sef-report, many were achieving, with some significanty exceeding, the recommendations. This is supported by the finding that non-participants were found to be more active on the GPPAQ. Of those citing medica reasons, it was ess cear whether or not these probems constituted a definite contraindication, especiay as those with predefined medica conditions contraindicating an increase in waking shoud have been excuded. A sma number of peope suggested that they did not enjoy PA, were not interested in waking or suggested a different activity or a team sport. 58 NIHR Journas Library

91 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 The externa theme reated to factors externa to the potentia participant, incuding work and other commitments, trave probems, being a carer and advice from others. Work and work-reated trave were frequenty given as reasons for not participating and many feared that they coud not commit to the tria. Famiy and home ife commitments were aso important barriers to increasing activity, incuding being a fu-time carer for famiy members. We were interested in finding out whether or not advice from friends or famiy affected the decision not to participate. Very few interviewees discussed participation; for those who did, it did not infuence their decision, except for one interviewee whose daughter strongy agreed that she shoud decine. Reasons reated to tria design incuded programme ength, tria materia, equipment probems, being randomised, the venue, the waking environment, the nurse interaction and the overa tria design. For some interviewees, the tria duration, at 3 months, was too ong and it was difficut to commit for this period. One interviewee reported a previous negative experience with pedometers as her primary reason for decining. Severa fet that not being abe to choose their aocated group was a disadvantage. Two interviewees expressed concern about the tria promoting waking as an exercise, because the oca waking environment was boring, and another stated that it was the wrong season for waking outdoors; however, most interviewees thought that waking was an appropriate and incusive activity. Some expressed interest in a group intervention rather than one to one with a nurse, feeing that this woud improve motivation and sociabiity; however, others fet that interacting directy with a nurse was preferabe to being in a group. Most interviewees approved of the choice of their GP surgery as the ocation for a PA intervention, describing their surgery as ovey, peasant, convenient and appropriate. Many interviewees expressed a positive attitude towards PA and research, and regretted not being abe to participate. Summary of the findings The PACE-UP tria recruited 11% of patients aged years who were invited by post from their registered genera practice, athough not a were randomised because of faiure to provide adequate acceerometry data. Those not repying were younger and more ikey to be mae and from deprived postcode areas. Asian patients were ess ikey to participate. Participation was associated with mid or moderate heath impairment, athough those with more severe probems were ess ikey to participate. Not having enough time and being aready physicay active were the most common reasons for non-participation, even among patients who were cassified as not active. Interview findings supported the questionnaire findings and gave more detai about the reasons behind the ack of time (work, trave, famiy, caring commitments, etc.) and the type of heath impairments that stopped peope from taking part. Despite not wanting to participate, amost a interviewees were positive about the tria, aware of the benefits of PA and the importance of research, and supported primary care as a venue for such programmes. The design of the tria and intervention was not stated as a key reason for decining to participate. Strengths and imitations The PACE-UP tria is a arge tria recruiting from a ceary defined invited popuation, based on GP ists, enabing us to assess the potentia reach of the intervention in terms of age, sex and deprivation. Our estimate of 11% participation may underestimate the true rate, particuary in areas of high mobiity where patients may have moved away and not informed the practice, infating the number of patients counted as invited. Athough based on imited data, the PACE-UP tria offers a rare opportunity to examine demographic differences between participants and non-participants. We were abe to estimate participation within different ethnicities using pooed data. However, we were not abe to match at an individua eve and some participants may have categorised themseves in a different ethnic group to that on the GP register. Ethnicity was aso poory recorded in some practices and we needed to make assumptions about whether or not those with unknown ethnicity were simiar to those with recorded ethnicity. In a sensitivity anaysis, even under extreme assumptions, the difference between Asian and back ethnic groups persisted. The tria excuded individuas who sef-reported being active, but the non-participants were not seected in this way. Our quantitative anaysis attempted to mitigate this difference by restricting anaysis to a those who sef-reported as not being active. However, some residua bias may remain. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 59

92 GENERALISABILITY The interview study represents an innovative attempt to systemicay expore the reasons for non-participation with a purposive sampe of those who were invited but decined. Our aim was to further understand the perceived barriers to participation to enhance recruitment to future trias and exercise programmes. We were aso abe to expore non-participants perception of the tria design and research in genera. This sampe spanned six out of seven of the practices invoved and incuded both sexes, a range of ages, ethnicities, empoyment and educationa backgrounds. The teephone interviews aowed in-depth exporation of the barriers to participation not possibe from a questionnaire aone and aowed us to compare the interview findings with the non-participant questionnaire responses. The main imitation of the interview study is that this was based on a sef-seected group of peope who both returned the non-participant questionnaire and agreed to be interviewed. It is possibe that some of our sampe woud have been excuded in any case on the basis of their pre-existing activity eves and, therefore, their decision to decine may have been entirey appropriate. In addition, despite our attempt to sampe interviewees from non-white British backgrounds, these groups were under-represented when compared with the ethnic diversity of the popuation invited. Comparisons with previous work A systematic review of 47 studies of waking interventions 151 showed that recruitment methods and participation rates were poory reported. Of 25 RCTs, participation rates coud be cacuated for ony five. We recruited by post to reduce the burden on practice staff and to obtain response rate data. Posta invitations are used in primary care for other preventative activities, making this a pragmatic approach. 152 Other waking intervention trias using posta invitations in primary care 33,48,144,153 had simiar response rates of 10 20%; those with higher rates (37% and 39%) 154,155 recruited individuas who were oder and more frequent attenders in primary care, and invited individuas in the primary care consutation, as we as by post. Our previous tria 21 used simiar recruitment strategies to the PACE-UP tria, but, unike other trias, incuded active peope. This tria had a recruitment rate of 30%, but was conducted among oder peope in an affuent setting with few non-white residents. Non-responders were foowed up with one reminder etter, but because of data protection constraints we were unabe to contact patients by teephone. Athough ony 1% of invitation etters were returned to sender, this may underestimate those who did not receive the etter as we did not use registered post. A previous London study using registered post found that 26% of etters were not deivered. 156 Our findings of greater participation in women, 157,158 oder peope 158 and those in affuent areas 157 are supported by other studies. Attwood et a. 158 found no association with deprivation or ethnicity, but this tria was based in an area of high deprivation with few non-white patients. Among Asian patients, our response rate was simiar to posta invitations in the PODOSA (Prevention of Diabetes and Obesity in South Asians) tria (5.2%), in which community-based approaches, 159 through partnership with oca South Asian groups, were found to be more effective. Wibur et a. 160 found that socia networking was the most effective method for recruiting African American women from ow-income areas. The finding that decining participation in this PA tria was attributabe to interviewees considering themseves to be aready sufficienty active is in ine with other iterature. 157, Importanty, objective measurement of PA reveas that most peope overestimate their activity eves, 9 and that their assessment of their persona activity eves is ikey to be infuenced by a socia context. 161 However, this interview series aowed activity eves to be expored in more detai and reveaed that, at east on sef-report, this was a reativey active cohort for some of whom the tria may not have been appropriate. Our finding that participation was associated with some degree of heath probems, but that severe impairment reduced participation, is more nuanced than previous work, which has suggested that decining participation in PA programmes or trias is attributabe to medica probems, incuding pain, 157,161,162,164 particuary in studies with oder participants. 165 Lack of time because of work and other commitments has aso been identified as an important barrier, 157, ,166,167 particuary in younger and midde-aged participants. 165 A ack of interest in PA has aso been reported in the iterature, 157, ,166 but trave away from home has not been reported prominenty. This may refect the high proportion of our interviewees sti in fu- or part-time work and the seasona migration of the diverse south-west London popuation. 60 NIHR Journas Library

93 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Impications Guideines pubished by NICE 36 concuded that more research was needed to determine which interventions are effective and cost-effective in increasing activity eves among ower socioeconomic and high-risk groups, and that there is itte evidence on differentia effects of interventions. In our tria, those groups for which more evidence is required tended to be those with the owest recruitment rates, such as Asians and those in more deprived areas. 143 It has been suggested 168 that specific cutura groups may respond better to interventions directy targeted at their needs, rather than to universa interventions. Reasons for non-participation were often reated to individuas not wanting to increase activity or feeing that they were sufficienty active. It is ikey that such resistance wi simiary appy to any intervention ro-out and may appy more widey to other pubic heath interventions. Low participation rates mean that poicy-makers shoud be cautious about the intervention s potentia reach and the possibiity that it coud increase activity inequaities, but these are not a reason not to impement an intervention shown to be effective in 11% 93 of the popuation. We were successfu in recruiting oder peope, women and those with comorbidities or some degree of heath imitation. These groups have ower PA eves and are ikey to benefit more from increased PA. However, those with more severe disabiity, peope who have had fas and those with a fear of faing were not over-represented, indicating a rationa choice by individuas. Ony 12% of non-participants cited randomisation as a factor for not participating, whereas 45% cited time constraints. The nurse intervention required three additiona visits to the practice on top of the three data coection visits, which may deter working peope or those with chid-care and other commitments. However, the PACE-UP tria showed that both the nurse-support and posta deivery groups performed simiary at the main 12-month outcome. 93 An intervention offering pedometers with brief advice, without the need to provide research data, may be more acceptabe. Both the PACE-LIFT and the PACE-UP trias recruited to target, achieved foow-up rates of over 90% and demonstrated that the interventions were effective in increasing PA eves. 21,93 However, consideraby more research effort was required (e.g. more contacts from research assistants) per randomised participant in the PACE-UP tria than in the PACE-LIFT tria, resuting from a ower uptake rate. In spite of the effort, we sti had imited power to investigate ethnic and socioeconomic subgroups. Trias with greater reach are ikey to be more expensive in terms of recruitment, and gains in generaisabiity need to be baanced with greater costs. Our findings have important impications for those panning PA trias, as we as for those commissioning community PA programmes. As the cohort we interviewed appeared to be reativey physicay active, it may be necessary to taior some interventions to maintaining, rather than increasing, activity; this may aso be important to mitigate the decrease in PA eves that often occurs with ageing. Equay, education about the eves of activity that optimise heath gain may prevent potentia participants from decining because of overestimation of their actua eves of activity. Measurements using pedometers or acceerometers provide an easy approach to vaidating PA eves. Lack of time was an important barrier, so it may be hepfu to reiterate that activity can be broken up into 10-minute bouts throughout the day (this can aso hep those individuas imited by pain or disabiity). Taioring interventions to an individua s trave and work commitments and for their specific heath probems may aso increase uptake. Promotiona materia shoud be incusive and expicity state that pre-existing medica conditions do not necessariy prevent participation and dispe myths about the risks of moderate-intensity exercise. Information about the vaue of PA, particuary waking, for many different heath conditions 1 shoud be emphasised in the invitation to participate. RCTs inevitaby invove randomisation, but emphasising that, in some trias (incuding the PACE-UP tria), the contro group can receive the intervention after the tria may hep recruitment. Most interviewees fet that primary care was an appropriate, convenient ocation for deivering a waking-based PA intervention, indicating that further primary care-based trias and programmes are ikey to be we received. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 61

94 GENERALISABILITY Concusions Participation in an effective PA tria among aduts and oder aduts in a sociay and ethnicay diverse popuation was ony 11%, with ower rates in more deprived and Asian subgroups, imiting the tria s abiity to investigate differentia effects in these important subgroups. Trias with greater reach are ikey to be more expensive in terms of recruitment, and gains in generaisabiity need to be baanced with greater costs. Differentia uptake of interventions found to be successfu in trias may increase inequaities in PA eves and shoud be monitored. 62 NIHR Journas Library

95 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 6 Process evauation Introduction Why is process evauation necessary in the PACE-UP tria? The PACE-UP RCT is a compex intervention comprising mutipe interacting components (pedometer, handbook, diary, practice nurse PA consutations and BCTs as part of both written materias and consutations). Athough the RCT design is abe to estabish the effectiveness of the intervention (see Chapter 3), it does not provide information on how it works, or whether or not there are contextua factors that coud be associated with variation in outcome in different settings. 169 Conducting a process evauation of the PACE-UP tria enabes a detaied examination of the mechanisms of change by gaining an understanding of how the intervention was deivered and received and how this may have affected the variation in outcomes. The process evauation investigates the reationship between the fideity and quaity of impementation, the context of the intervention and the main tria outcomes. The evauation aso heps to draw concusions on the repicabiity and generaisabiity of the intervention. The main findings from the process evauation are pubished 170 and are reproduced here under the terms of the Creative Commons Attribution License ( Medica Research Counci s guidance on process evauations in compex interventions In 2014, the Medica Research Counci (MRC) pubished new guidance for process evauations of compex interventions. 169 The guidance draws on the experiences of researchers and wider stakehoders who have conducted process evauations within trias of compex pubic heath interventions. We have used the guidance to provide the framework for the process evauation of the PACE-UP tria. Process evauation is accompished through investigating aspects such as impementation, mechanisms of impact and context, and the reations between these, as described in Figure 11. Impementation refers to the structures, resources and methods through which deivery is reaised, and comprises the foowing factors: impementation process, reach, fideity, dose and adaptations. The impementation process describes how the deivery is achieved, through training, support, resources, etc. Reach refers to coverage and the degree to which the intervention is deivered to those for whom it was intended, that is, who receives the intervention. The other aspects of impementation are reated to what is deivered. Fideity is the degree to which the intervention was deivered as intended (content) and incudes assessment of the quaity of the intervention. Dose denotes the quantity of the intervention impemented. Adaptations are participant and impementer adjustments, which may impede or strengthen the intervention and which arise in response to the intervention itsef. Mechanisms of impact refer to how the intervention activities and participants responses to them cause change and adaptations. Context refers to externa factors which may infuence, and be infuenced by, impementation mechanisms and outcomes. The first stage of designing the process evauation was to describe the intervention and to carify casua assumptions, which was accompished through the use of a ogic mode, shown in Figure 12. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 63

96 64 NIHR Journas Library Context Contextua factors that shape theories of how the intervention works Contextua factors that affect (and may be affected by) impementation, intervention mechanisms and outcomes Causa mechanisms present within the context that act to sustain the status quo or enhance effects PROCESS EVALUATION Description of intervention and its causa assumptions Impementation How deivery is achieved impementation process training, resources Who it is deivered to reach What is deivered Fideity Dose Adaptations Mechanisms of impact Participant responses to and interactions with the intervention Mediators Unanticipated pathways and consequences Outcomes FIGURE 11 The key functions of process evauation and the reationships between these. Green boxes represent the components of process evauation, which are informed by the causa assumptions of the intervention, and inform the interpretation of outcomes. Adapted from Moore et a. 169 This is an Open Access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, for commercia use, provided the origina work is propery cited. See:

97 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Probem Evidence base Resources Activities Short-term outcomes Most aduts and oder aduts ead sedentary ifestyes and do not do enough PA for heath benefits Being physicay active reduces the risk of overa mortaity and prevents or reduces compications from > 20 heath conditions Increasing waking can safey increase moderateintensity PA Pedometers and step count goas can hep aduts and oder aduts to increase waking BCTs hep peope to change behaviour and maintain changes Grant funding Seven genera practices and practice nurses Mutidiscipinary research team incuding GPs, epidemioogists statisticians, heath economists, psychoogists, pubic heath, PA and quaitative experts, BCT trainers, tria manager, research assistants Equipment (pedometers, acceerometers) Patient resources (diaries, handbooks, etc.) Nurse training in BCTs, PA promotion, pedometers, safety-reporting Nurse-support group: 12-week programme, three PA consutations with nurse, pedometer, handbook, diary Posta deivery group: 12-week programme, pedometer, handbook, diary Contro group: offered nursesupported or posta intervention after 12 months foow-up Assessment of PA and other outcomes at baseine and 12 months (face to face) and 3 months (posta) FIGURE 12 Logic mode for the PACE-UP (Pedometers and Consutation Evauation UP) PA tria. Trained nurses demonstrate competence and confidence in PA consutations Participants in both intervention groups use pedometers to record steps in diaries and work towards individua waking targets Medium-term 3-month (interim) outcomes Acceerometrymeasured change in step count, time in MVPA, time spent being sedentary from baseine Change in patient-reported outcomes from baseine: heathreated quaity of ife, depression and anxiety scores, pain, exercise sef-efficacy Increase in time spent waking change in waking habits Longer-term 12-month (main) outcomes Acceerometrymeasured change in step count, time in MVPA, time spent being sedentary from baseine Change in patient-reported outcomes from baseine: heathreated quaity of ife, depression and anxiety scores, pain, exercise sef-efficacy Anthropometric changes since baseine (BMI, waist circumference, body fat) Increase in time spent waking change in waking habits and onger-term heath benefits DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO

98 PROCESS EVALUATION Methods and resuts The PACE UP tria process evauation was conducted aongside the effectiveness evauation, incuded both quaitative and quantitative components and was undertaken by the same team that carried out the effectiveness evauation. In accordance with the MRC guidance, the methods were seected through foowing the key functions mode (see Figure 11) and are summarised in Tabe 16, which detais the process evauation components, the data sources, the tria group to which they refer and the measures used. The nurse-supported intervention was the most compex to deiver as it invoved eight nurses from seven practices deivering three consutations over a 3-month period. Most of the process evauation was therefore designed to evauate the nurse-supported intervention group. When process evauation occurred for other groups, this is ceary described. TABLE 16 Summary of the PACE-UP tria process evauation data sources, evauative groups and reported measures Process evauation component Data source Tria groups evauated Measures Impementation Impementation process How was it deivered? Training: Nurse training day agendas BCT trainer teephone feedback records (generated from audio-recordings of nurse intervention sessions) Tria administrative records Nurse-support group Time spent on training activities Resources: Tria administrative records (see Chapter 4) Nurse-support and posta deivery groups Cost of deivering intervention components Reach Who was it deivered to? Tria recruitment records (see Chapter 2) Data coection on non-responders and non-participants, incuding non-participant interviews (see Chapter 5) A groups and non-participants Recruitment frequencies and percentages Quaitative themes and subthemes from non-participant interviews Fideity (content and quaity) What was deivered? Nurse session checkists (see Appendix 5) Patient aiance questionnaire (see Appendix 5) Nurse aiance questionnaire (see Appendix 5) BCT trainer feedback sheets (generated from audio-recordings of nurse intervention sessions) Nurse-support group Content number of items deivered (mean and SD) Quaity of deivery (frequencies and percentages) BCT competency scores, measures of quaity of deivery (mean, SD and range) PA diaries Participant interviews and nursesupport focus groups and interviews (see Chapter 7) Pedometer use questionnaire (see Appendix 5) Nurse-support and posta deivery groups Competed diary return and weeky target achievement (frequencies and percentages) Participant and nurse quotations, quaitative themes and subthemes Pedometer use (frequencies and percentages) Dose What was deivered? Nurse session checkists (see Appendix 5) Audio-recordings of nurse intervention sessions (used to generate BCT trainer feedback sheets) Nurse-support group Sessions attended (frequencies and percentages) Consutation durations (mean SDs, medians and interquartie ranges) 66 NIHR Journas Library

99 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 16 Summary of the PACE-UP tria process evauation data sources, evauative groups and reported measures (continued ) Process evauation component Data source Tria groups evauated Measures Adaptations What was deivered? Nurse training session records PA diaries Participant interviews and nurse focus groups and interviews (see Chapter 7) Nurse-support group Nurse-support and posta deivery groups Comments made by nurses Ateration of targets (frequencies and percentages) Participant and nurse quotations Mechanisms of impact Participant responsiveness Patient aiance questionnaire (see Appendix 5) Nurse aiance questionnaire (see Appendix 5) Nurse-support group Measures of responsiveness (frequencies and percentages) Participant and nurse quotations Participant interviews and nurse focus groups and interviews (see Chapter 7) Nurse-support and posta deivery groups Quaitative themes and subthemes Context Contextua factors Nurse training session records Participant interviews and nurse focus groups and interviews (see Chapter 7) Nurse-support group Nurse-support and posta deivery groups Comments made by nurses Quaitative themes and subthemes Participant and nurse quotations To reduce dupication, and for ease of reading, the methods and resuts for each aspect of the process evauation are presented together. The main resuts are summarised in a further key functions mode after a of the resuts have been presented (Figure 13). Severa aspects of the process evauation are deat with appropriatey in other chapters of this report, such as reach in Chapter 5 and participant responsiveness in Chapter 7. They are referred to in Tabe 16 for competeness and the reevant chapter in which the methods and resuts for this aspect are presented are ceary shown. We have seected three quantitative aspects of the process evauation to reate directy to PA outcome measurements at 12 months (change in step counts and change in time in MVPA in bouts): the number of nurse appointments attended whether or not competed step count diaries were returned by participants at 3 months in the nurse-support and posta deivery groups, and the use of pedometers in the nurse-support and posta deivery groups during the 12-week intervention. These anayses reating process to outcome measures are described at the end of Methods and resuts. Impementation Impementation process Training Nurses deivering the intervention were provided with training in PA guidance (1 hour and 25 minutes), tria protocos (4 hours), safety reporting (1 hour and 10 minutes) and BCTs (9 hours and 25 minutes) across the duration of the tria (see Appendix 5, Tabe 38). Data on nurse training were obtained from training-day agendas and BCT trainer teephone feedback records and tria administrative records. The tota training time was approximatey 16 hours; most of this was aocated to deivering BCTs as the active ingredient in the intervention. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 67

100 68 NIHR Journas Library Description of intervention and its causa assumptions Primary care pedometer-based 12-week waking intervention, with and without practice nurse support (posta or nurse supported by three PA consutations) Assumption that providing a pedometer, individua waking target and handbook based on behaviour change techniques wi increase PA eves Context Factors affecting (and affected by) impementation of this waking intervention: season / weather (effect of rain, ice and snow on waking), environment (easier to wak in nearby park than in buit-up area), heath issues (waking making pain worse and pain improved by waking) and empoyment (retired peope having more time for waking, some occupations providing waking opportunities), observance of reigious events (difficuty achieving PA targets during Ramadan and Christmas) and socia factors (waking with partner/friend/famiy / grandchidren, or not having anyone to wak with). Contextua factors often ed to adaptations by nurses or participants Impementation Process training 16 hours mainy BCTs, aso PA, protoco and safety Reach 10% of those invited participated in the tria Fideity content we covered. 80% returned competed PA diaries Fideity quaity 90% of participants fet understood and respected. Nurses were competent, proficient or expert at BCT deivery Dose 74% attended a three nurse sessions Adaptations (often reated to context) for heath imitations, pain, weather, reigious observances, hoidays and coupes Mechanisms of impact Participant responsiveness 90% of participants fet that the pedometer used in the PACE-UP tria was hepfu 83% of participants fet that the number of nurse appointments was just right PACE-UP has changed my ife... Outcomes 12-month outcomes Significant increases in average daiy step count and average weeky MVPA in bouts in nurse-support and posta deivery groups compared with contro (no difference between intervention groups) Nurse-support group increased sef-efficacy compared with contro No change in other outcomes Significant associations between impementation measures (diary return, pedometer use and nurse sessions attended) and PA outcome measures PROCESS EVALUATION FIGURE 13 The key functions of process evauation and the reationships between them for the PACE-UP tria. Green boxes represent components of process evauation, which are informed by the causa assumptions of the intervention and inform the interpretation of outcomes. Adapted from Moore et a. 169 This is an Open Access artice distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) icense, which permits others to distribute, remix, adapt and buid upon this work, for commercia use, provided the origina work is propery cited. See:

101 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Resources Resources for the tria incude the tria materias [patient handbooks and diaries, see the NIHR Journas Library website: (accessed 25 May 2018)], tria equipment (pedometers and acceerometers) and payment of nurse time and room hire. These are a fuy costed in Chapter 4 and are not further commented on in this chapter. Reach The overa tria recruitment rate was 10% (1023/10,467). Detais on how practices and participants were seected and recruited is described in Chapter 2. The methods for assessing tria representativeness and generaisabiity, by comparison of non-responders, non-participants and participants, and by interviews of non-participants, are described in Chapter 5. Fideity (content and quaity) Content Nurse sessions Nurse session attendance and session content deivered was recorded by the nurses after each session (see Appendix 5). There were 11 compusory items to be covered in session 1 and six items to be covered in sessions 2 and 3. The eve of nurse session attendance was high; approximatey three-quarters of participants attended a three sessions (n = 255/346; 74%). Adherence to content deivered was high in a sessions; the mean number of items deivered in each session was 11 (range 10 11) and six (range 5 6) in sessions 2 and 3, respectivey. Of those participants who attended session 3, most reported sti using the pedometer and diary (n = 258/263; 98%) (Appendix 5, Tabe 39). Physica activity diaries Physica activity diaries (see Appendix 5, Tabe 40) returned by participants after the intervention provided data on the achievement of weeky waking targets for the intervention groups. Eighty per cent of participants (n = 549) returned competed diaries; there was simiar return across both groups. One-third of participants in the nurse-support group atered their step count targets (89/346; 32%) and the majority were decreased (n = 80). In comparison, just four participants in the posta deivery group atered their step count targets and a were decreased. The reationship between diary return at 3 months and the tria outcome measures was expored (see the association between the process evauation measures and the tria outcomes at the end of Methods and resuts). Pedometer use A participants were asked about their pedometer use during the 12-week intervention period (see Appendix 5, Tabe 41). During the 12-week intervention, a high proportion of both the posta deivery and nurse-support groups reported using their pedometer either every day or most days: 238 out of 294 (81%) in the posta deivery group and 269 out of 303 (89%) in the nurse-support group. The reationships between pedometer use during the intervention and the tria PA outcomes at 3 and 12 months were expored for both intervention groups (see the association between the process evauation measures and the tria outcomes at the end of Methods and resuts). Quaity Nurse and patient aiance questionnaire Foowing the intervention, both the nurse and the participant independenty competed a 12-item nurse and patient aiance questionnaire (see Appendix 5) covering different intervention aspects (e.g. working together and goa-setting, number of appointments). The questionnaires were deveoped by BCT trainers and tria investigators, and questions 1 11 were adapted from the Working Aiance Inventory, 171,172 which is a vaidated aiance measure used in cognitive behavioura therapy-based studies, and the Outcome Rating Scae. 173 Item 12 was added to specificay ask about the number of PACE-UP tria appointments. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 69

102 PROCESS EVALUATION The questionnaire was posted to the participant and returned to the researcher, so that the nurse was bind to participant responses. Three directy comparabe items (questions 1, 3 and 4) from both the patient questionnaire and the nurse questionnaire provided data on the quaity of intervention deivery (Tabe 17). Seven directy comparabe items directy reate to participant responsiveness (questions 5, 6, 8, 9, 10, 11 and 12). Two items (questions 2 and 7) were discounted as they did not reate to quaity of participant responsiveness. The questionnaires were competed by 295 out of 346 participants (85%) in the nurse-supported intervention group and by the nurses for 251 out of 346 nurse-support group participants (73%). There was strong agreement between the participant and nurse resuts for a of the items reating to quaity and 80% or more of both nurses and participants agreed or strongy agreed with a of these statements, suggesting high quaity of deivery. TABLE 17 Quaity of deivery and participant responsiveness data from the nurse and patient aiance questionnaires Questionnaire Patient aiance Nurse aiance Deivery and responsiveness N Agree or strongy agree, n (%) Missing items N Agree or strongy agree, n (%) Missing items Quaity of deivery Q1: the patient and I worked together on setting goas that were important to the patient Q3: the patient fet heard, understood and respected Q4: in our meetings together, the patient discussed everything they wanted to discuss (80) (88) (90) (94) (94) (94) 2 Participant responsiveness Q5: the patient understands how to make asting changes in activity eves Q6: the approach to making change suited the patient Q8: the patient fees confident to continue to make positive changes in PA on their own Q9: the patient fees confident about overcoming obstaces to increasing activity eves in the future Q10: the pedometer used in the PACE-UP study was hepfu to the patient Q11: the diary used in the PACE-UP study was hepfu to the patient (90) (86) (86) (73) (83) (78) (66) (65) (90) (85) (81) (83) 4 Q12: the number of appointments with the PA nurse was just right (83) (74) 10 Notes Missing items were excuded from the percentage cacuations. Q9 on the nurse aiance questionnaire was printed bank on the Likert scae for answers, so a high number of responses were missing. 70 NIHR Journas Library

103 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 The foowing are some exampes of participant and nurse comments reating to quaity from the questionnaires: Nurse was encouraging, supportive. Encouraged me to set goas that were achievabe for me and not to put too much pressure on mysef. Femae, aged 47 years, nurse-support group My nurse was ovey and encouraged me a the way through, even when some days I coudn t do what I needed, we discussed aternatives. My nurse was a diamond. Thanks to PACE-UP and the nurse my waking has reay improved. Femae, aged 47 years, nurse-support group Cient peased with programme. Learning curve. Woud recommend to others. Enjoyed patient, good discussions and understanding around increasing exercise. Practice nurse Practice nurse Audio-recordings from nurse sessions Nurses were asked to audio-record a sampe of their sessions so that these coud be istened to by the BCT trainer and rated according to their ski in six different communication ski competencies. Ratings were made by the BCT trainer against a primary care consutation rating scae (range 0 6) in six domains (Figure 14) and used for both fideity (quaity) evauation and supervision purposes. The rating scae used was deveoped from the Cognitive Behavioura Therapy Techniques for Paiative Care Practitioners Rating Scae 174 and the Department of Heath and Socia Care s The Ten Essentia Shared Capabiities A Framework for the Whoe of the Menta Heath Workforce. 175 The nurses were each asked to provide three audio-recorded sessions, one each for sessions 1, 2 and 3. They were asked to try to ensure that one of the recorded sessions was from a session where a coupe were seen together. The mean scores and ranges for a nurses are shown across a domains in Tabe 18. The range of scores iustrates that even the owest ratings were competent, and the highest scores were expert, across a six competencies. The owest scoring competency was given for empowering expanations (mean score 4.7), whereas a the other competencies had mean scores above 5, demonstrating proficiency, with very good features. Competence eve Incompetent Exampes 0 Absence of feature, or highy inappropriate performance 1 Inappropriate performance, major probems evident Novice 2 Evidence of competence, but numerous probems and ack of consistency Advanced beginner 3 Competent, but some probems and/or inconsistencies Competent 4 Good features, but minor probems or inconsistencies Proficient 5 Very good features, minima probems and/or inconsistencies Expert 6 Exceent performance, or very good even in the face of patient difficuties FIGURE 14 Behaviour change technique competency eve. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 71

104 PROCESS EVALUATION TABLE 18 Fideity: quaity scores of performance for audio-recordings of nurse sessions by BCT trainer Communication ski competency Scores 1. Framing, pacing, focus and use of time 2. Empowering expanations 3. Coaboration and active istening; interpersona effectiveness 4. Setting goas, agreeing actions and motivationa techniques 5. Feedback, reviewing and summarising 6. Buiding sef-efficacy Average (mean) score Range of scores Quaitative perspective Semistructured individua interviews with participants and focus groups with nurses provided a quaitative perspective of the intervention, incuding the quaity of deivery; this is presented in detai in Chapter 7. Overa, the nurses and participants described the intervention in a positive manner, as highighted by the foowing quotations:... they kept saying how we I was doing, and a this sort of thing, so it made me want to continue. I think it was... a part motivation, yes, because I knew I had to face somebody and I didn t want to fai. Femae, aged 63 years, nurse-support group... if you had, in your drawer, you had ike a set... a package, programme, you coud do, and if through the NHS Heath Check you identified someone who was suitabe, you coud then discuss it with them and say, Woud this be something you d be wanting to ook at?... and go from there. Practice nurse Dose For the purpose of the PACE-UP tria, the dose deivered to the posta deivery group was fixed, as they a received the same handbook, diary and pedometer. The dose coud vary for the nurse group according to the number of sessions attended and the ength of each of the sessions. Nurses were required to compete checkists at the end of each session (see Appendix 5, Tabe 39), providing detais on attendance and the duration of sessions. The duration of sessions was aso captured from the audio-recorded sampe of intervention sessions; this aowed a comparison with session durations cacuated from nurse checkists. Overa, three-quarters of participants in the nurse-support group attended a three sessions. Ninety-five per cent of participants (330/346) attended session 1, 86% (296/346) attended session 2 and 76% (263/346) attended session 3. The reationship between the number of sessions attended and tria PA outcomes was aso expored (see the association between process evauation measures and tria outcomes at the end of Methods and resuts). Tria protocos detaied the foowing approximate duration of each nurse intervention session: 30 minutes for session 1 and 20 minutes each for sessions 2 and 3. A summary of nurse intervention session durations from nurse sef-report and audio-recordings is avaiabe in Appendix 5, Tabe 42 (19 recordings, reating to 22 participants, as some were coupes). There was good agreement between the panned protoco session ength and the nurse sef-report durations: mean of 30 minutes (SD 4 minutes) for session 1, mean of 24 minutes (SD 3 minutes) for session 2 and mean of 22 minutes (SD 4 minutes) for session 3. The duration of consutations from audio-recordings was based on much smaer numbers (n = 22 participants) but had a shorter mean duration: mean of 21 minutes (SD 6 minutes) for session 1, mean of 21 minutes (SD 7 minutes) for session 2 and mean of 14 minutes (SD 5 minutes) for session NIHR Journas Library

105 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Adaptations Detais of nurse and participant adaptations made to the intervention were provided from nurse training session records. There were many exampes presented reating to step count target adaptation and taioring the intervention to individua circumstances. Adjustments were made to the intervention to accommodate reigious observances, such as Ramadan and Christmas. Step count targets were adapted to be more achievabe to refect participants reduced energy/activity eves in advance of hoidays, when there were expected to be reductions or increases in PA, during periods of participant iness and pain and in response to changing weather conditions. Nurses aso expained the need for fexibiity with participants who experienced difficuties with equipment use; for exampe, a sma number of participants who did not ike using the pedometer were advised to use time to measure their waking, rather than measuring step count (e.g. extra waks of minutes per day, rather than an extra steps per day). At the second training session, it became cear that the nurses did not find the optiona handouts provided for use in consutations to suppement the patient handbook hepfu and, as a consequence, were not using them. From these discussions, it was decided that these materias woud be discontinued. Another adaptation reveaed by the nurses was adapting targets and advice for participants taking part as a coupe, particuary if they had very different PA eves and targets; this sometimes reated to one individua of the coupe having heath probems that had an impact on mobiity. Nurses adapted the intervention to encourage both participants to meet their targets, often encouraging a mixture of waking together and waking apart to achieve this. Physica activity diaries aso provided data on adaptions to the prescribed intervention through aterations of participants waking targets (see Appendix 5, Tabe 40). Few participants in the posta intervention group atered their targets in the diary (1%; 4/339), whereas 32% (89/346) of the nurse-supported intervention group atered their waking targets, mainy by decreasing the target [29% (80/346)]. Additiona detais of nurse and participant adaptations to the intervention by both intervention groups, obtained as part of the quaitative evauations of the tria, are presented in Chapter 7. Some exampes are given here. The nurse quotations beow iustrate the fexibiity in intervention deivery during Ramadan and aso during bad weather:... they coudn t wak or increase on their waking at that time because they hadn t eaten and then they weren t feeing too good, and a that, so we did it a different way then, and what I did with them was we reaxed it and then I said to them, when Ramadan s over, we made the appointment so that their actua tria went on a bit onger. Practice nurse... if the weather was bad, or it was cod, or there were obstaces that got in the way... they woud do things ike, you know... activities indoors where they coud not aways go outside. Practice nurse Mechanisms of impact Participant responsiveness Seven items refecting participant responsiveness were identified in both the patient aiance questionnaire and the nurse aiance questionnaire (see Tabe 17). There was a good degree of consistency between participant and nurse responses for a of the items reating to participant responsiveness and high eves of agreement with a of the statements. For exampe, 90% of participants said that the pedometer was hepfu and 83% said that the number of nurse appointments was just right, suggesting that there was a good eve of participant responsiveness to the intervention. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 73

106 PROCESS EVALUATION Some exampes of both participant and nurse comments reating to participant responsiveness from the aiance questionnaires are shown beow: PACE-UP has changed my ife. I use the car ess when I go about. Athough I drive to work I park about 1 km away from work, then wak a the way to and from. Femae, aged 47 years, nurse-support group Wearing the pedometer reay raised my awareness of how far I waked each day. I wi continue to use it. Femae, aged 68 years, nurse-support group Positive, iked study book. Was a very refective/honest person regarding his exercise. Positives/ negatives easiy identified by patient. Practice nurse Patient has own exce monitor of readings/pedometer count. Aso converts to mies daiy. Practice nurse Information about participant responsiveness aso came from the quaitative evauations of participants from both the posta deivery and the nurse groups (from individua interviews) and from practice nurses (from a focus group and individua interviews), and is presented in Chapter 7. The foowing quotations from the quaitative evauation iustrate participant responsiveness and engagement from a nurse and participant perspective:... the ony other thing I d say about the diary is that the peope that reay iked fiing it in found it a reay good motivator. When they came to the ast appointment, they wanted another one. Practice nurse There s nothing ike the fact that you know you re going to meeting someone and tak about it to make you do it, you know,... It s basicay the routine of being checked up on by someone ese... Mae, aged 61 years, nurse-support group... we having something which counts the steps makes one conscious of it and fiing out a itte booket every day, ikewise, it just creates some persona pressure. Mae, aged 59 years, posta deivery group Context Comments made by nurses during training sessions reevant to contextua factors were noted down. There was overap with the factors mentioned in the section on adaptation of the intervention, as contextua factors often required the nurses to consider adapting the intervention or targets after discussion with participants. Exampes of contextua factors mentioned are as foows: the difficuty of waking in bad weather, the effect of taking part in the intervention as a coupe, heath issues that required a sower, more gradua approach and undertaking the intervention during Ramadan or Christmas. How contextua factors may have affected (and been affected by) the impementation, such as season, environment, heath status, empoyment and socia and reigious factors, were expored as part of the quaitative evauation of participants (from individua interviews) and nurse perspectives (from focus group and individua interviews), and are described more fuy in Chapter 7. The foowing factors were described: season/weather (probems with rain or snow and ice), environment (ease of waking in parks, more difficut in buit-up areas), heath issues (exampes both of pain getting worse with waking and of waking improving 74 NIHR Journas Library

107 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 pain), empoyment (retired peope saying they have increased time for waking, some occupations providing the opportunity for waking or at east waking to work), reigious factors (difficuty with waking during Ramadan when fasting and having ow energy eves) and socia factors (waking with famiy/grandchidren, etc. or not having anyone to wak with). This ast factor is iustrated by the foowing quotation taken from the quaitative evauation:...it s something I want to keep up, because I just fet that it was such a benefit, and even the kids woud come out with me sometimes. Nurse-support group participant Association between process evauation measures and tria outcome measures Athough the tria was powered ony for anaysis of the difference in outcome measures between the three groups, and not for exporation of the effect of process evauation measures, we fet that it was interesting to expore if there was any reationship between adherence to the intervention and the change in outcomes. We have focused on three quantitative measures of process evauation in reation to the PA outcome measures at 3 months and 12 months (changes in average daiy step count and weeky time in MVPA in bouts). The three measures were a to do with the impementation of the intervention: 1. Dose nurse session attendance (0, 1, 2 or 3 sessions attended; nurse-support group). 2. Fideity return of competed diaries after the 3-month intervention (yes/no; posta deivery and nurse-support groups). 3. Fideity pedometer use how often did you wear the pedometer? Every day or most days (yes/no; posta deivery and nurse-support groups) during the 12-week intervention (0 3 months). A measures described were considered as independent variabes in the modes, with (1) change in average daiy step count and (2) change in tota weeky MVPA in bouts as outcomes. A anayses were adjusted for age, sex, practice, month of baseine acceerometry, househod identifier (to account for custering by househod) and tria group (as in the main tria anayses; see Chapter 2). Tabe 19 shows the resuts of the modes. Nurse session attendance and physica activity outcomes In the nurse group at 3 months and 12 months there was a positive association between the number of sessions attended and the PA outcomes. Participants attending a three sessions increased their step count and their time in MVPA in bouts at 3 and 12 months by significanty more than those attending between 0 and 2 sessions. Diary return and physica activity outcomes In the posta deivery group, there was a strong positive association between returning a diary and on both change in steps and MVPA in bouts at both 3 and 12 months compared with those in the posta deivery group, who did not return a diary. In the nurse-support group, there was a positive association between returning the diary and on change in step count and MVPA at 3 months, compared with those in the nurse-support group who did not return a diary. However, by 12 months, there was no significant association for either PA outcome of returning a diary within the nurse-support group. Pedometer use and physica activity outcomes In the posta deivery group, reported use of a pedometer every day or most days during the 12-week intervention period was associated with a significant change in step count at 3 months and 12 months, and with change in MVPA at 3 months (borderine effect at 12 months). Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 75

108 PROCESS EVALUATION TABLE 19 The PACE-UP tria modeing resuts: reating nurse session attendance, step count diary return and pedometer use to PA outcomes PA outcome Daiy step count Tota weeky minutes of MVPA in 10-minute bouts 3 months 12 months 3 months 12 months Intervention components Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue Nurse session attendance Attended a three nurse sessions: yes vs. no 1197 (627 to 1766) < (74 to 1137) (45 to 103) < (3 to 57) 0.03 Diary returned Posta deivery group: yes vs. no 1458 (854 to 2061) < (538 to 1689) < (33 to 94) < (17 to 75) Nurse-support group: yes vs. no 873 (190 to 1555) ( 278 to 925) (15 to 85) < ( 27 to 33) 0.89 Pedometer use every day or most days during the 12-week intervention Posta deivery group: yes vs. no 1029 (383 to 1675) < (22 to 1190) (6 to 73) ( 2 to 55) 0.07 Nurse-support group: yes vs. no 337 ( 525 to 1198) ( 321 to 1109) ( 20 to 68) ( 25 to 45) Notes A modes incude practice, sex, age at randomisation and month of baseine acceerometry as fixed effects and househod as a random effect in a mutieve mode Within the nurse-support group, there were no significant associations between reguar pedometer use during the 12-week intervention and change in step count or MVPA at 3 months and 12 months. This ack of significant effect coud be expained by the very sma numbers of participants in the nurse-support group who reported not having used a pedometer reguary during the 12-week intervention (n = 34; 11%). Overa, the anaysis of the association between process measures and PA tria outcomes exhibits a cear pattern of positive associations (i.e. increased nurse appointments, diary return and pedometer use were associated with increased objective PA eves). This provides cear evidence of the engagement with the tria process and outcomes, but cannot be interpreted as causaity. Discussion Main findings Figure 13 summarises the key findings from the PACE-UP tria process evauation, which foowed the MRC guidance for process evauation of compex interventions. 169 We gathered a number of positive data on impementation, suggesting good-quaity intervention deivery and adherence to the protoco, despite the ow reach of the tria. Nurse training was an important eement of the tria, with nurses receiving approximatey 16 hours of training, predominanty around deivery of BCTs. We demonstrated good 76 NIHR Journas Library

109 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 coverage of the proposed session content by the nurses and aso good-quaity deivery, with the audio-recording of nurse sessions demonstrating high eves of competency in communication skis. High-quaity deivery was aso refected in comments from participants who fet heard, understood and respected. Three-quarters of the nurse group attended a three PA consutations, and around 80% of both the nurse-support and the posta deivery groups engaged with the sef-monitoring aspects of the tria and returned competed step count diaries at 3 months. In terms of the mechanisms of impact, we demonstrated high eves of participant responsiveness. Context was important and factors affecting the impementation of the waking intervention were suggested by nurses and participants from both intervention arms, and incuded the effects of weather, the environment, heath issues, pain, empoyment, observance of reigious events and socia factors. The nurses worked with participants to hep them to make adaptations, either to the intervention or their individua targets, or to encourage participants to come up with soutions when possibe, when there were contextua chaenges. Severa process evauation measures (number of nurse sessions attended, return of a competed diary and reguar pedometer use) showed significant associations with PA outcomes at 3 and 12 months. Strengths and imitations Strengths Despite the PACE-UP tria process evauation having been designed before the pubication of the MRC guidance, 169 we were abe to use the data that we coected and fit them into the framework, which has provided a usefu structure for reporting process evauation findings. We have a number of different data sources that refect three different perspectives in the tria: participant, practitioner and observer. This provided a broader picture of the process than many studies have reported and aowed comparison of resuts from different methods (e.g. for consutation duration). We tried to reduce the burden of participant measurement, tria costs and dupication of effort, by coecting as many routine tria data as we coud for the evauation (from administrative records, nurse training records, nurse checkists, nurse audio-recordings from supervision sessions, etc.), but we suppemented this with data coected specificay for the evauation (e.g. the nurse and patient aiance questionnaires or the 12-month pedometer use questionnaire). We have used a mixed-methods approach to the process evauation, as recommended, combining quantitative data from key process variabes from a participants with in-depth quaitative data from purposivey seected sampes. 169 The quaitative eement is described in fu in Chapter 7; however, we have used reevant quotations in this chapter to iustrate the quaity of deivery, adaptations, participant responsiveness and context, and these have provided a voice to participants and nurses and added a richness and depth to the evauation. The data were coected ongitudinay and contemporaneousy throughout the tria, which is seen to be the most compete and accurate method of data coection, and aso aows any change in intervention deivery over the course of the tria to be detected. 169 The data are comprehensive, with a high response rate and competeness of data sources, strengthening the robustness of the findings. The process evauation anaysis was conducted before the outcome anaysis to avoid a biased interpretation of the process data. Ony the fina anayses examining the effects of process evauation variabes on outcome data were carried out foowing the main outcome anayses. Limitations The process evauation was conducted by the tria team whie the tria was ongoing. This aowed efficient data coection in a contemporaneous manner, but coud have ed to bias in evauation. We tried to minimise the bias by using objective instruments when possibe (e.g. nurse and patient aiance questionnaires, 12-month pedometer use questionnaires, return of patient diaries). In addition, the quaitative evauation was ed by Christina Victor, who was not invoved in the day-to-day tria conduct. Some process measures were not fied out by everyone (patient aiance questionnaire: 85% competion; nurse aiance questionnaire: 73% competion); this coud have ed to a more positive assessment of statements, if those who fet negativey about the programme did not repy. Nurses coud have seected more positive consutations to audio-record, infating the BCT competency eves, athough they were encouraged to record cases as they occurred. Not a of the nurses submitted audio-recorded consutations with coupes, meaning that we were unabe to ook separatey at the quaity of these consutations. The content deivery for the nurse consutations was evauated Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 77

110 PROCESS EVALUATION from checkists fied out by the nurses because they may have overestimated what they had achieved. We tried to compensate for this by coecting additiona data from both the participant and observer perspective to hep to corroborate these data; both participant and observer data suggested a high degree of quaity in the consutations. The consutation durations were shorter from observer data than from sef-report, but data on observed consutations were based on a much smaer sampe. The study was not powered to ook at the effects of adherence to different aspects of the protoco on tria outcomes; we therefore have reduced power for these anayses, which imits the interpretation of the findings, which cannot be taken to be casua. Comparison with other compex intervention process evauations A number of studies have examined intervention impementation fideity, with a arge variety of process structures and methods; therefore, it is difficut to draw direct comparisons. Process evauations have become increasingy important, but the purposes and design of studies have been mixed. Many process evauations are competed independenty of tria data coection and are observations of a random subsampe of participants or practitioners. 176,177 The process evauation of the PACE-UP intervention provides both participant and nurse perspectives and identifies a ink between contextua factors and adaptations in intervention deivery and acceptance. The study aowed us to ook at both perceived and observed behaviour change in the nurse intervention deivery and participant responsiveness; this is unike many other studies, which have tended to focus on ony one perspective, which is most often the person deivering the intervention. 176, Nurse comments coected at training, individua session checkists and nurse and participant comments from the quaitative work iustrate the intervention deivery and adaptations in context. The intervention was designed to bring about change at an individua eve when deivered in a primary care environment; we have observed that context infuences the deivery and impementation of the intervention through adaptation. This is simiar to findings from other PA and dietary compex intervention studies with process evauation. 182,183 Specificay, Fitzgerad et a. 182 identified that negotiation and fexibiity pay an integra roe in overcoming the barriers and resistance to change in a dietary intervention. Previous studies have coected data at an organisationa or practice eve; 176,181 there are few studies that have captured the evauation of behaviour change at an individua eve and from two perspectives. A previous study that did ook at both patient and practitioner perspectives, however, reported much greater variations in dose and adherence to protocos than seen in the PACE-UP tria, therefore making it difficut to estabish which eements of the intervention were effective. 184 Berendsen et a. 184 reported that many heath-care professionas deviated from the protoco of a ifestye intervention to accommodate individuas and reduce faout, which was associated with increased patient satisfaction for the intervention sessions. This perhaps suggests that adaptations and taioring of an intervention have a strong infuence on retention, adherence and, possiby, effectiveness in ifestye and behaviour change interventions. The PACE-UP intervention aowed nurses to adapt the sessions as necessary to each individua, whie maintaining the key deiverabes in each nurse-ed session; athough we have not ooked at the effect of adaptation on retention, we have seen that dose (nurse session attendance) was associated with effectiveness of the intervention. This, in turn, promotes the consideration of buiding adaptations and fexibiity into intervention design. Our finding of an association between the return of a competed step count diary and a change in PA outcomes is consistent with the findings of a systematic review, 31 which suggested that the use of a step count diary was common to many successfu pedometer interventions. Impications of the process evauation for the interpretation of the PACE-UP tria We have demonstrated that the PACE-UP tria had good adherence to the protoco, the intervention was acceptabe and was rated positivey by both the nurse-support group participants and the posta deivery group participants, and both groups engaged in sef-monitoring using the pedometer and step count diary. It is not possibe to infer causaity directy from the process evauation data, but the high eve of engagement with pedometers and diaries by both intervention groups suggests that these were important factors in heping peope to make the PA changes observed. This is supported by the associations demonstrated between increased PA eves and the foowing process measures: number of consutations attended, return of a competed step count diary and pedometer use. The carefu description and documentation of the tria processes, the coection of additiona data for the process evauation and the pubication of the resources used as appendices mean that our intervention and process evauation woud 78 NIHR Journas Library

111 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 be easy for others to repicate, from training through deivery, to foow-up and evauation. Use of the MRC framework gave a ogica and coherent structure for reporting, 169 which is aso easy for others to foow. The PACE-UP tria had a positive and significant effect on PA outcomes, but had this not been the case, the positive process evauation with high eves of fideity woud have enabed us to have confidence that any negative tria effect woud not have been because of poor tria impementation. The tria demonstrated a stronger effect on the main PA outcomes and on exercise sef-efficacy at 3 months in the nurse-support group than in the posta deivery group, athough the effects on PA outcomes were simiar between the groups at 12 months (sef-efficacy remained higher at 12 months in the nurse-support group). The process evauation demonstrated that the nurses were deivering BCTs in their PA consutations in accordance with the protoco and with high eves of competence (in addition to the BCTs provided in the handbook and diary for both groups). This suggests that the nurse-deivered BCT eements of the intervention have strong short-term effects on PA eves (and, possiby, onger-term effects on sef-efficacy). The possibe effects on onger-term maintenance are examined in Chapter 8. The impications of the ow reach of the tria for generaisabiity and the pubic heath impact have been discussed in Chapter 5 and wi be considered further in the main discussion in Chapter 9. The process evauation demonstrated important contextua factors that had an impact on participants abiity to engage with a waking intervention; these shoud be considered in any future ro-out of the programme, particuary regarding how the programme may need adaptations to be made in these circumstances. Concusion The PACE-UP tria process evauation demonstrated that the tria was we deivered by the tria team and we received by participants. The MRC framework was a usefu vehice for reporting the evauation. An association between adherence to the tria protoco and main tria PA outcomes has been demonstrated. Important contextua factors were shown that may need adaptations to be considered in any ro-out of the intervention. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 79

112

113 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 7 What did the nurses and participants think about the intervention? Introduction It is important, within the deivery of a behaviour change intervention tria, to understand the experiences of those invoved in the deivery and receipt of the intervention. In order to address these important issues, two quaitative studies were embedded within the tria protoco. In these studies, we sought to gain insights into two important issues: (1) the views and experiences of the nurses deivering the intervention and (2) the experiences of tria participants. Summaries of the nurses perspectives on the tria 185 and the participants evauation of the tria in heping them to increase their eves of PA 162 have been pubished and are reproduced here under the terms of the Creative Commons Attribution 4.0 Internationa License ( In this chapter, we focus upon the perceptions of the nurses and participants of the tria; some of the quotations have been pubished previousy in the pubications detaied above. It is of fundamenta importance to understand the perspectives of both those invoved in the deivery of the intervention and those who received the intervention. In Chapter 6, we provided a comprehensive process evauation of the tria which described, in detai, the training and support given to the tria nurses. In Chapter 6, we aso addressed issues reated to the fideity of the tria deivery by the nurses. In this chapter, we focus on the nurses perspective on their experience of participating in the tria, with a view to understanding how we can better pan and deiver primary care-based trias and then impement them more widey. There is an extensive iterature that examines adherence to behaviour change interventions in aduts and which estabishes the barriers to, and faciitators of, for exampe, increasing PA. For exampe, Picorei et a., 186 focusing on oder aduts, reported that adherence to exercise interventions was associated with key demographic factors (higher socioeconomic status and not iving aone), heath status (fewer heath conditions, taking fewer medications and better sef-rated heath) and psychoogica factors (fewer depressive symptoms). 186 For tria participants, a nove part of our quaitative study addressed a reated, but much ess we researched, issue of trying to understand why the intervention did, or did not, work as a means of evauating the individua eements of the intervention in faciitating behaviour change. Recruiting tria participants for the quaitative study At initia recruitment into the tria, participants were asked for consent to participate in foow-up teephone interviews, such as those incuded in this aspect of the study. The tria statistician prepared a spreadsheet of a participants who had competed the 12-month foow-up in January 2014 and who had given consent to be approached to participate in the teephone interviews (this ist was updated in March 2014). We purposey recruited participants who had, and had not, increased their PA eves from both nurse-supported and pedometer-ony intervention groups. We defined an increase as 200 steps per day; anyone who either did not achieve this or decreased their PA was defined as a non-improver. This gave us four interview groups: (1) nurse support/increase, (2) nurse support/no increase, (3) posta deivery/ increase and (4) posta deivery/no increase. We aso ensured that we samped participants with a range of ages, from both sexes and from a six of the initia participating practices. As noted in Chapter 2, a nove feature of our tria was the option for participants to take part as a coupe and we wanted to ensure their incusion in our quaitative study. We purposey targeted potentia participants from demographic groups under-represented in our main sampe (e.g. ethnic minority community participants) to ensure that we expored the widest range of views possibe. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 81

114 WHAT DID THE NURSES AND PARTICIPANTS THINK ABOUT THE INTERVENTION? Between February and Apri 2014, we identified 96 tria participants who had been seected on the basis of the criteria described above. We made contact with 44 of these participants, of whom ony one decined to be interviewed. We were unabe to make contact after three attempts with the remaining 52 participants, who had been initiay identified as potentia participants. The 43 participants we recruited broady approximated to the demographic parameters specified, with 20 participants in the 45- to 59-year-od age group, 29 participants being femae, 21 participants in the nurse-supported intervention group and 23 participants who did not increase their step counts. We interviewed seven participants who took part as a coupe, but we did not interview both partners. In terms of ethnicity, 29 were white British and a further five were from other white ethnicity groups, with nine participants from back and Asian ethnic minority groups. Appendix 6, Tabe 43 shows the demographic and step count detais of individua participants. We used a semistructured interview guide taiored for each group to refect the nature of the interventions and the cassification of participants as improvers/non-improvers. 162 The 43 interviews competed ranged in duration from 9 to 44 minutes, with an average (mean) duration of 21 minutes. Interviews were recorded and transcribed verbatim. Fu detais of the anaysis strategy for our quaitative interviews is provided esewhere, 162 but are briefy summarised here. A transcripts were read by four authors (RN, JS, CV and TH), codes were assigned independenty and discrepancies were resoved by discussion. Codes were grouped into themes, which were further refined by discussion to produce broader themes, encompassing severa subthemes. Theoreticay informed BCTs were an important eement of the tria and we were interested in understanding which of these had been of most use to participants. We performed an additiona anaysis of the data to specificay draw out themes reevant to these techniques. We have reported the reasons for tria non-participation and the barriers to, and faciitators of, increasing PA esewhere. 148,162 We have noted esewhere that, athough we defined our groups by the quantitative increase in waking, responses from participants did not demonstrate this distinction. In this, and in our previous quaitative evauation from the PACE-LIFT study, 149 amost a participants interviewed fet that they had benefited from the trias, even if this had not been manifested by an increase in their step count. In this chapter, we focus on what participants tod us about their motivation for participating in the tria, their experiences of the various components of the tria and the onger-term impact of tria participation. The roe of nurses in the PACE-UP intervention As described in detai in Chapter 6, the tria intervention was deivered by eight nurses across seven practices. They deivered three PA consutations to participants in their arm of the tria. These took pace in weeks 1, 5 and 9 of the 12-week pedometer-based waking programme. Participants deveoped an individuaised PA pan with the practice nurses, based around their current eve of activity, with the goa of increasing both step count and time spent in MVPA. The nurses provided each participant with an individua PA diary, incuding step count targets for the 12 weeks, based on their own baseine PA measures, but this coud be taiored further in the nurse PA consutations through joint discussions between nurse and participant. Five PACE-UP tria nurses participated in a focus group that was ed by two of the research team (CB and CV); additionay, two nurses were interviewed individuay by Rebecca Normanse (who aso attended the focus group) and one further nurse was not avaiabe to participate in this phase of the research. A further focus group was aso carried out with nurses invoved in the previous PACE-LIFT tria of a pedometer-based waking intervention with oder peope. 21 The focus groups asted for, on average, 106 minutes and the individua interviews asted for 50 minutes. The pubished evauation of nurse experiences of the interventions incudes data from both trias. 185 In this chapter, we have imited the resuts presented to those from the PACE-UP tria nurses. A semistructured interview guide was used to eicit the nurses views on their participation in the tria (see Appendix 6). The interviews/focus groups were audio-recorded and transcribed verbatim. Fu detais of the methods used are avaiabe esewhere, 185 but are briefy summarised here. Coding the transcript themes was guided by thematic anaysis for both the group and individua interviews, and areas of disagreement were discussed to ensure a consensus. Researchers were mindfu that group interviews refect a generaised understanding, whereas individua interviews provide more persona views. However, simiar interpretations and themes emerged from both types of interview, and referra 82 NIHR Journas Library

115 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 to fied notes throughout the process enhanced the trustworthiness of the findings through data trianguation. 185 In the rest of this chapter, we combine the perspectives of both the nurses and the tria participants to provide an overview of their experiences of being part of the PACE-UP tria. We present our resuts in terms of the three key phases of the intervention: preparing for the tria, deivering or receiving the intervention and after the tria/impementation. Preparing for the tria A key theme that emerged from our work with the nurses was the importance of the pre-tria training programme. As documented in Chapter 6, before the tria was impemented the nurses received training in the BCTs that underpinned the intervention. Support was then ongoing once the tria had started. Athough these were experienced practice nurses, we coud not presume detaied famiiarity with a of the techniques that our study invoved. An additiona and important feature of the training was the importance of tria fideity. One unique eement of our tria was the deivery to coupes. This is not a famiiar service deivery mode for our nurses and so it was one area where we provided specific support/training. As comprehensivey demonstrated in Chapter 6, the consensus from our nurses was that they fet appropriatey trained and prepared to deiver the intervention as per the protoco: a the training was reay exceent (PACE-UP nurse, focus group). For participants, the key pre-tria activity focused upon the decision to participate. This is described in detai in Chapter 5, in which we discuss the recruitment and participation rates. In Chapter 5, and in Normanse et a., 148 we present both the sociodemographic profie of participants and examine why individuas opted not to participate in the tria. Athough we have detais of the sociodemographic profies of those who take part in PA trias, we have ess information on what motivated participants to take part in the tria. This was not expicity expored in our interview, as the focus was on the tria and its impact on eves of PA. However, some participants taked expicity about their reasons for taking part in the tria. Key individuay based motivations to participate were concerns about weight and heping to manage existing ong-term conditions (especiay diabetes meitus). Others noted the commitment they were making in signing up for the intervention, even though they were not aware of which group they were in when they agreed to participate in the tria: We I must admit, when I first signed up for it, I was thinking what I have done to commit mysef to this for quite a ong time (ID30). Deivering and participating in the tria Our focus in this chapter is on the experiences of participating in the tria, rather than on the outcomes or the factors that faciitated changes in PA. Our tria had three arms: usua care (n = 338), posta pedometer intervention (n = 339) and nurse-supported pedometer intervention (n = 346). Of our 43 interviewees, 21 were in the nurse-supported intervention group and 22 were in the posta deivery group. We first consider the experiences of participants who were in the nurse-support group and compare these with the posta deivery group in terms of how they perceived their engagement with the tria. Perceived vaue of nurse consutations The nurse-ed consutations were we accepted by the participants in that part of the tria, with 74% (255/346) attending a three sessions. Overa, those who were aocated to the nurse-supported intervention arm were very happy with their meetings with the nurse, as iustrated thus: Yes, the nurse was very hepfu... it was reay good for me. ID12 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 83

116 WHAT DID THE NURSES AND PARTICIPANTS THINK ABOUT THE INTERVENTION? Conversey, those who had not been in the nurse-support group were, overa, confident that they did not need the support of the nurse: I think I was happy doing it on my own. I don t think... I mean it was very... it was very easy to foow your instructions and what you wanted us to do and so I don t think meeting a nurse... ID13 No, I don t think it [visiting the nurse] woud have been usefu reay. ID17 Some were sceptica of what advice the nurse coud give: No, not reay. What woud she say, wak a bit quicker, eat a bit ess? It s common sense, I knew in the first pace. ID23 I know what I shoud be doing. ID8 Others were confident in their own interna motivation: I think if I ve agreed to do something, then I wi try and achieve that target, whether somebody tes me face to face or by post, so I think it s dependent on the individuas maybe, individua choice. ID38 There were two key caveats to the confidence of the posta deivery group in increasing their PA without the support of a nurse. These were existing heath probems and overcoming barriers. Severa participants in the posta deivery group observed that if they had had a ong-standing heath probem, they may have preferred the security and support of the nurse, as iustrated in this comment: If I did have heath probems, I may have wanted to see a nurse, and say, ook, I ve been doing this and I ve had an ache and a pain here, sha I stop or... you know, if it was that situation, and I think yes, you might need to speak to someone you know who coud advise you medicay, but I didn t need it. ID13 In addition, some participants in the posta deivery group thought that being abe to see the nurse might have heped them to overcome the difficuties and chaenges they experienced when trying to increase their PA, for exampe: I woud have found that better [to see a nurse] because, if I d have taked to her about the steps, she might have been abe to umm introduce something ese... [participant was concerned that the target step count was unachievabe]. I think if I d have been seeing the nurse reguary then, during that summer, umm, we woud have found another way I fee, you know. ID19 Behaviour change techniques Our intervention incuded over 20 distinct behaviour change activities, as defined by Michie et a., 68 embedded within the PACE-UP tria handbook and diary (received by both the posta deivery group and the nurse-support group) and, additionay, within the protoco for the nurse consutations. Reference to these specific BCTs were then extracted from the interview transcripts to determine what eements of behaviour change were viewed as being most hepfu by the participants. 162 There were 152 exampes of these factors in our interviews: 54 in the posta deivery group and 98 in the nurse-support group. With the exception of sef-monitoring, the BCTs were more frequenty noted in the nurse-support group 84 NIHR Journas Library

117 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 compared with the posta deivery group. The eements of Michie et a. s 68 typoogy that were especiay evident in our participants comments were (1) the provision of information, (2) monitoring and feedback and (3) strategies for reapse prevention/overcoming chaenges. Rewards, an important eement of the typoogy, were not seen as being of great importance by our participants. Both of our intervention groups appreciated receiving information about the ink between behaviour and heath. However, a key and important type of information provided, especiay in the nurse-support group, was specificay taiored and personaised information about how, where and when to increase waking, for exampe: She gave me a printout of umm... some... waks that you coud do, group waks and things ike that. ID18 In terms of monitoring and feedback, nurses payed an important roe for their group, as exempified by comments such as the foowing: She was very encouraging. ID39 Oh I fet reay happy [with the nurse] and she was very happy too with me and I did reay ike my nurses, yes. That was one of my reasons because, each week I go, I ask her if they think I am doing we, am doing we, yes [aughs]. ID12 Importanty, the nurses were seen to provide motivation and encouragement when the novety of the intervention was waning and participants were at risk of apsing, for exampe: I think there was a point where they sort of said, you know, don t give up now, or something ike that, you know, at the point where... the novety might have worn off... ID35 The posta deivery group argey fet that they coud sef-monitor their activity: I fet ike it was enough to know [the step count]... I think it was fine, just to sort of keep... I was very good about fiing the diary in and it was sort of for me that was enough to keep me going reay. ID43 Adapting the tria protoco versus fideity Given that this was a tria, it was essentia that the nurses deivered the intervention as per the protoco; therefore, before the tria started, we provided extensive training to the nurses and emphasised the importance of adherence to the prescribed protoco: The equipment was exceent, the pedometers, the acceerometers, exceent, exceent, exceent. However, given that the tria ran over 12 months, pragmatic adaptations were made by the nurses (in consutation with the team) in response to the specific circumstances of participants, for exampe, working around hoiday periods (e.g. Christmas) or periods of reigious observance, such as Ramadan:... forget about the 2 months around Christmas... you can t get appointments and they don t want to wear it [pedometer]. This serves to remind us that deivering behaviour change interventions in rea-word practice requires fine tuning in the deivery to refect the compexity of peope s daiy ives. We assessed fideity by audio-recording a minimum of three consutations for each nurse, as described in detai in Chapter 6. However, this aso enabed us to provide specific feedback to the nurses on their use of the BCTs empoyed. This was universay Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 85

118 WHAT DID THE NURSES AND PARTICIPANTS THINK ABOUT THE INTERVENTION? wecomed and improved their practice during the tria, but aso provided them with enhanced skis to take into practice beyond and after the tria: I actuay changed my practice from thereon, so yes, it was exceptionay hepfu. Tria materias and equipment Both nurses and participants remarked on the materias that supported the tria, namey a pedometer, a PA diary and a series of optiona handouts, as described in Chapter 2. From the participants perspective, each eement of the tria had both advantages and disadvantages. However, the focus of the comments about equipment from the participants perspective was upon the accuracy (or otherwise) of the pedometer. Typica of the more critica comments of the pedometer was: That day we went for a reay ong wak round the common, I was reay disappointed when I come back, it didn t seem to register very many steps. I ve obviousy done a ot more than... registered about,000 or something, we waked around the whoe of Tooting, is a bit more than that I think. ID7 There were fewer comments about the diary, which was seen as being motivating: I was very good about fiing the diary in and it was sort of for me that was enough to keep me going reay. ID43 However, for others it was chore: Like I say, it is an effort, it s umm... you have to know what you are in for, and then reay maintain it. I had to record it every day, yes, you are busy, sometimes you are out and you go for dinner and then you come home and then you had a few drinks and you can t remember what day it is. ID1 Actuay writing down the activities and things, it... after about the first day, I got bored with that. ID3 Nurse satisfaction One feature of the tria noted by the nurses (but not the participants) was the time that they had to devote to the PA consutations compared with their norma activities, as this comment iustrates: You know, we don t have any protected time for heath promotion... the heath promotion is the add-on. It s giving us the time, because we don t have the time. From the nurses perspective, this engagement provided considerabe job satisfaction in seeing their participants embrace change and become more physicay active. A feature not experienced prior to the tria by most of the nurses was deivering the intervention to coupes rather than individuas. This presented a unique chaenge for some of the nurses as, in norma practice, it is unusua to be working with two patients simutaneousy. Sometimes the dynamic worked we:... most coupes, they enjoy doing it together because they d go... they coud go out waking together and, even if it was through the winter, at east if they were both going, they had each other... they use to encourage each other. So if one didn t want to go the other one woud encourage them and they d make sure they went. 86 NIHR Journas Library

119 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 At other times, the probems encountered were a significant barrier: I m not actuay overy sure how coupes worked. I don t know if I had, I don t know if it caused more issues sometimes, in the fact with the pedometers, because they got so focused sometimes on the fact that their pedometers didn t match up. After the tria and impementation A participants fet that they had benefited, regardess of changes in their objective PA eves, and had deveoped skis in terms of embedding PA in their daiy ives and routines and by deveoping strategies to overcome chaenges when they arose: Yes, everyone in my house now, we don t drive to the shops, we a wak to the shops... it was easier for me just to jump in to the car, now I have to think twice, do I reay have to? Yes, setting my own targets and now... we, it s something that I ve got used to now and I m determined to keep it up. ID21 ID11 A of our nurses were very positive about how participation in the tria had deveoped and enhanced their knowedge and skis, which they were appying across a wide range of routine ifestye consutations, not just reated to PA, but aso for smoking cessation, weight oss management and the prevention of chronic diseases. From the perspective of primary care and the nurses specificay, participation in the tria generated a egacy and the project was, in a sma way, abe to support the deveopment of expertise in primary care of the use of BCTs and in working with patients in different ways (e.g. coupes). Our intervention was an individua as opposed to a group-based intervention. Some participants firmy beieved that this was most appropriate: If it invoved each person reporting back on their success or faiure at meeting the previous targets, it might be a bit awkward in a group possiby. However, others identified the potentia benefits of a group intervention: When you are with other peope, and then you see the same probems they are facing, some of them might come up with other ideas... you can form a team, support network. ID21 Utimatey, this suggests that we need a repertoire of interventions that mesh with the circumstances and preferences of the different popuations. There was virtuay uniform support for the ocation of the tria within a primary care context, with many participants recognising the convenience of getting to their GP surgery: ID2... you woudn t want someone to have to trave and peope know how to get to their doctors don t they? Yes, that was good, because obviousy it was very near home so it was idea. ID 29 ID15 Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 87

120 WHAT DID THE NURSES AND PARTICIPANTS THINK ABOUT THE INTERVENTION? From the nurses perspective, athough acknowedging that the intervention woud be beneficia to their patients, they observed that within the time constraints of routine practice, they woud not be abe to repicate the fu intervention as it stood within a routine nurse consutation. The nurses made suggestions for modifying the intervention to focus on the pedometer and printed materias for use opportunisticay within norma practice and, perhaps, to be avaiabe on prescription. The suggestion that heath advisors, or a reated roe, coud deiver the intervention that we evauated was not supported by our nurses. Discussion From this phase of our study, severa key points arise. Amost without exception, both the nurses and the participants enjoyed taking part in the study and fet that it had provided them with important and enduring benefits. For the nurses, these benefits were couched in terms of new skis that they coud transfer into their routine practice. For participants, there was an increase in their awareness of the benefits of waking as a means of enhancing heath by increasing PA. Among our intervention groups, this perceived benefit was articuated irrespective of the objective changes in PA. For nurses and intervention group participants, a key feature of the tria was the preparatory work before the intervention started. This was especiay important for the practices who participated and the nurses. We opted to deiver this tria in ordinary genera practice settings to test out the potentia of the intervention to be impemented in routine primary care settings. Participants and nurses aike fet that primary care is the appropriate setting from which to run such interventions. However, the nurses provided a caveat to this with comments about the constraints of time within rea-word primary care. Prior to the impementation of the intervention, the research team and our behaviour change experts worked extensivey with the practices and nurses and provided training in the intervention, feedback on the deivery of the intervention and support across the tria to the nurses deivering the intervention. One exampe of a chaenge with which some nurses needed support was in working with coupes. This is not something that usuay occurs within their genera working regime and support from the research team in deaing with these chaenges was important. The nurses were supported to adapt how materias were introduced and used within the consutations to make these materias reevant to the participant and thereby personaise the intervention more. This is a key chaenge in effectivey deivering standardised interventions, in both tria settings and every day primary care how to ensure the consistency of information provision and support, whie aso making it reevant to individuas. Empowering nurses, and other primary care staff, to make patient-centred adaptations to standard behaviour change programmes is ikey to resut in improved outcomes. This inks to the important issue of taioring support to make changes in heath behaviours to match the circumstances and preferences of individuas. Thus, support needs to be appropriate to externa factors, such as seasons or the time of year, or key events in peopes ives, such as retirement, but aso refect individua circumstances such as preferences for group or individua activities and the reevance of written or other types of digita materias and current heath probems. Our study has incuded two types of intervention that offered varying eves of support, both of which generated increased eves of PA. A key chaenge for future studies is to determine which groups woud benefit from the minima support pedometer by post-type intervention, and those for whom the more intense nurse-ed intervention is the most appropriate. 88 NIHR Journas Library

121 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 8 Three-year foow-up to assess the maintenance of physica activity eves Introduction The PACE-UP anayses showed positive effects on 12-month PA eves (see Chapter 3). We wanted to see if this effect was maintained at 3 years, as this has important impications for the NHS; specificay, woud any future pedometer programme require a top-up after 12 months? Our interventions ed to an extra minutes weeky of MVPA in bouts (an increase of about one-third from baseine) in a predominanty inactive cohort. Deivery via three nurse PA consutations had the same effect on 12-month outcomes as the simper, cheaper posta deivery. These are exciting findings, as they show that a ow-cost posta pedometer intervention increases PA eves in sedentary aduts and oder aduts. However, it is vita to know whether 12-month effects persist at 3 years or if a further intervention boost is needed. We therefore successfuy appied for additiona funding from NIHR s Heath Technoogy Assessment programme to foow-up the tria cohort at 3 years (2 years after the previous 12-month foow-up) with both quantitative and quaitative evauations. Both the quaitative 187 and the quantitative findings 188 have been pubished and are reproduced here under the terms of the Creative Commons Attribution License ( To date, itte is known about the ong-term sustainabiity of PA interventions. A meta-anaysis of interventions (incuding pedometers) to increase PA eves in 55- to 70-year-ods incuded ony four trias with data beyond 12 months (a sef-reported). They found a imited evidence base beyond 12 months, and caed for more trias with a onger foow-up period and objective PA measures. 26 These findings were supported by a Cochrane systematic review 23 and recent NICE guidance on PA interventions. 29 The ProAct 65 + tria 189 of a PA intervention found that between-group differences persisted at 2 years post intervention, but ony for sef-reported PA. As we as a ack of ong-term objective PA data after interventions, there is aso a ack of quaitative evidence on maintenance. A iterature review suggested that PA disengagement usuay occurs 6 months after an intervention has ended, but caed for more research to distinguish the factors that ead to successfu and unsuccessfu PA maintenance. 190 A very sma primary care study foowed up participants 6 months after a pedometer-based intervention, and found some usefu insights to expain how this pedometer intervention worked and how it may be deveoped, 44 but further quaitative studies on onger-term effects are acking. The PACE-UP tria 3-year foow-up provides evidence on objective PA eves, 2 years after the 12-month foow-up. After the 12-month foow-up, 212 out of 322 contros (66%) received a pedometer, handbook and diary by post. They had no further input (unike the origina posta deivery group, who were teephoned by a research assistant 1 week after being sent the pedometer, to check that it had arrived, and who were asked to return their competed PA diaries for review at 3 months). The contro participants being sent the pedometer by post mimics what woud happen if this simpe, pragmatic intervention were to be roed out by post through routine primary care, without any further input. As described in our protoco, 70 a further 64 out of 322 contro participants (20%) attended a singe nurse appointment after the main tria, in which they were given a pedometer, diary and handbook, but, again, received no further contact. The other intervention groups received no further intervention after the 12-month foow-up [apart from a sma number of the posta deivery group, 50/312 (16%), who had a singe nurse PA appointment]. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 89

122 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS Anthropometric measures did not show differences at 12 months; therefore, data on these were not coected at the 3-year foow-up. Therefore, our foow-up study focused on estabishing evidence of effectiveness at 3 years in objective PA measures and had the foowing objectives: to investigate if the origina nurse-support and posta deivery groups showed any persistent intervention effect (change in step count and time in MVPA in bouts) at 3 years, compared with the baseine eves to investigate if there were any differences between the origina nurse-support and posta deivery groups in their change in objective PA eves (step count and time in MVPA in bouts) between baseine and 3 years to investigate if the simpe posta pedometer intervention at 12 months increased objective PA eves (step count and time in MVPA in bouts) in the contro group compared with baseine. We aso fet that it was important to expore how participants in the nurse-support and posta deivery groups fet about the interventions in terms of maintenance of any increase in PA eves, and what factors might hep to encourage this further or to overcome barriers they had to increasing or maintaining their PA eves. Furthermore, we were interested in how the initia contro group fet about the minimaist intervention that they received. Our quaitative evauation therefore had the foowing objectives: 1. quaitative evauation of both the nurse-support and posta deivery intervention groups to ook at factors affecting PA eves and maintenance of any increase in PA eves at 3 years 2. quaitative evauation of the contro group to see the effect of the minima intervention on PA eves. Methods for quantitative physica activity evauation The 3-year foow-up focused on coecting the objective PA acceerometry data and other questionnairebased sef-reported outcomes by post to minimise data coection costs. To aow for seasona variation in PA eves, the baseine, 12-month and 3-year outcomes needed to be assessed in the same caendar month; therefore, foow-up ran from October 2015 to November Ethics approva and research governance Ethics approva for the 3-year foow-up was granted to the tria from London, Hampstead Research Ethics Committee (reference number 12/LO/0219). NHS oca research and deveopment approva was granted to cover a of the practice sites. Participants eigibe for the 3-year foow-up A tria participants who had not withdrawn from the tria were eigibe to be foowed up, even if they had not provided 3- or 12-month foow-up data. Lists of eigibe participants were organised by practice and practices were asked to check whether any participants had died, moved away or deveoped a termina iness or dementia since tria participation. These patients were then excuded. Contacting participants Eigibe participants were contacted with a etter expaining the tria 3-year foow-up. A participant information sheet, consent form and a Freepost return enveope were aso incuded, as was information on the main 12-month tria resuts. The etter expained that a research assistant woud contact the participants by teephone in around 1 week s time to discuss the 3-year foow-up. If they were happy to take part without further discussion, they were invited to post back the signed consent form. Informed consent The research assistant contacted participants by teephone approximatey 1 week after sending out the etter about the 3-year foow-up to discuss any questions they might have after reading the participant information sheet. Part of the consent form incuded consent to contact participants for an interview to 90 NIHR Journas Library

123 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 discuss their current PA eves in more detai. If the participants were happy to proceed with the 3-year foow-up, they signed and dated the informed consent sheet and returned the top copy to us. Data coection Once the informed consent for foow-up was agreed, the research assistant arranged a time to post out the acceerometer (GT3X+) to participants to measure their current usua PA eves for 1 week. Instructions about how to wear the acceerometer were incuded (on a bet, over one hip) and participants were asked to wear it for 7 consecutive days, from getting up unti going to bed, as they had done previousy. A diary was aso provided to record what activities were done and for how ong. They were aso sent a heath and ifestye questionnaire to compete (see Appendix 7; this was simiar to that competed previousy) and a short questionnaire about sef-reported PA eves to compete (the 7-day PA questionnaire used previousy) after they had finished wearing the acceerometer. They were provided with a Freepost return enveope to send the acceerometer and both questionnaires back. If the acceerometer did not record at east 5 days with at east 540 minutes per day, participants were asked to rewear it for a further week. The set of recordings with the greatest number of days with at east 540 minutes per day was incuded in the anaysis. Once acceerometers with adequate data were received, participants were posted out a 10 gift voucher. Outcome measures The main outcome measures (a acceerometry) used to evauate the 3-year foow-up were as foows: 1. change in average daiy step count, measured over 7 days between baseine and 3 years 2. change in time spent weeky in MVPA in 10-minute bouts between baseine and 3 years 3. change in time spent sedentary weeky between baseine and 3 years. Athough patient-reported outcome measures were coected from the heath and ifestye questionnaire (e.g. depression, 75 anxiety, 75 quaity of ife, 76 sef-efficacy, 74 pain, 77 disabiity 80 ) and from the 7-day PA questionnaire (IPAQ 72 and GPPAQ 73 ), these have not been assessed further as part of this report. Acceerometer data reduction ActiGraph data were reduced, as described previousy in Chapter 2, for the main tria. Anaysis summary variabes were aso identica to those used in the main tria, described fuy in Chapter 2. Procedure for accounting for missing data Ony days with at east 540 minutes of registered time on the acceerometer on a given day were used. The main anaysis of effect incuded a subjects with at east 1 satisfactory day of recording at 3 years. Statistica methods Statistica methods for the anaysis of the 3-year foow-up are argey as described in Chapter 2 for the primary anaysis at 12 months. Average daiy step count at 3 years was computed from a random-effects mode, aowing for day of the week and day order of wearing the acceerometer as fixed effects and participant as a random effect. The average daiy step count at 3 years was then regressed on average daiy step count at baseine with treatment group, age, sex, practice and month of baseine acceerometry as fixed effects and househod as a random effect, in a mutieve mode. The same anayses were carried out for MVPA in 10-minute bouts and daiy minutes of sedentary time. The primary anayses used the 681 participants who provided acceerometry data at 3 years. Sensitivity anayses were carried out to assess the effect of missingness: 1. Mutipe imputation methods were used to impute outcome data for those missing at 3 years, assuming that outcomes were missing at random, conditiona on variabes in the mode. We used the Stata procedure mi impute. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 91

124 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS 2. Missing-not-at-random anayses were used when it was assumed that changes in the contro group from baseine to 3 years were missing at random, but the change in each of the intervention groups was ± 500 and ± 1000 steps from their missing-at-random estimate. For this anaysis, we used a mean score method, 191 which has been impemented in a Stata modue, rctmiss, avaiabe from Statistica Software Components at (accessed 25 May 2018). Methods for quaitative evauation Samping Participants consented to be contacted for a teephone interview at the same time as they consented to take part in the 3-year foow-up. In February 2016, the research assistant produced three ists of participants who had aready provided 3-year acceerometry data and who had given consent to be interviewed (one ist for each arm of the tria nurse support, posta deivery, contro). The tria statistician randomy sorted these three ists ready for the quaitative researchers to start approaching participants for interviews. The interviewers were binded to participants previous and current PA eves. One of the aims of the quaitative evauation was to expore the success, or otherwise, of the minimaist intervention provided to the contro group after the main tria; therefore, the foowing participants were excuded from this quaitative evauation: contro group participants who had attended a nurse consutation after the main tria contro group participants who opted not to receive the posta pedometer after the main tria. At 12 months, participants in the posta deivery group were aso offered a nurse consutation. Again, those who attended the nurse consutation were excuded from the samping procedure, so that a those samped from the posta deivery group had received just the posta intervention. The aim was to interview approximatey peope from each arm of the tria (45 60 peope in tota), but continuing further if required, unti saturation was reached. As two researchers (CB and CW) were interviewing participants, they met reguary to discuss the samping, interview schedue and any emerging themes. Interviews were conducted unti saturation of new information was reached. By ooking at the participants demographic information it was possibe to ensure that the study group incuded both maes and femaes and aso represented a range of different ages and ethnicities, to ensure that a wide range of views were expored. Recruitment and informed consent Participants were initiay contacted via e-mai to arrange an appropriate time to contact them for a teephone interview. Participants without an e-mai address were caed and the interview was either conducted then or arranged for ater. To assess response, a detaied record was kept of when each participant was contacted, incuding information on who agreed, who refused and who coud not be contacted. Charotte Wahich and Caroe Beighton conducted the guided interviews with participants using a topic guide (see Appendix 7). Before the interview commenced, the participants were reminded of their initia consent to be approached for an interview; if the participants were happy to go ahead, their consent was then sought for the interview to be audio-recorded. Once the recording had started, the quaitative researchers stated the participant s ID number to ensure confidentiaity and anonymity in the subsequent transcript. On interview competion, the participants were offered a 10 high-street gift voucher to thank them for their time. Transcribing Interviews were prompty transcribed verbatim by an externa source. Once the transcripts were received back, they were doube-checked against each audio-recording by the quaitative researchers. Transcripts were aso circuated to the research team to ensure consistency between the interviewers and to hep assess when theme saturation had been reached. 92 NIHR Journas Library

125 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Interview schedue The interview schedues were deveoped through discussions with Tess Harris, Charotte Wahich, Caroe Beighton, Christina Victor and Rebecca Normanse. Sighty different questions were used for participants in the intervention groups (posta deivery and nurse support) and participants in the contro group (see Appendix 7). For those in the intervention groups, the aim was to expore participants views about PA maintenance and whether or not a top-up intervention was required, whereas for those in the contro group, the aim was to expore their views about the minimaist intervention. The interview schedues were revised sighty during data coection to ensure that the questions were cear and to incude additiona questions to gain a better understanding of the participants experiences. Anaysis A verbatim transcripts were read repeatedy by Charotte Wahich and Caroe Beighton. Initia ine-by-ine coding was conducted independenty to assign conceptua ideas to important episodes within the data. Through discussion with Rebecca Normanse, Tess Harris and Christina Victor, any discrepancies were resoved; this heped to ensure that the interpretation and categorisation of the data were vaid. After further discussions between Charotte Wahich and Caroe Beighton, these codes were then refined and grouped into emergent and anticipated themes. Resuts for the quantitative physica activity evauation Foow-up rate Of the 1023 origina tria participants, 32 had withdrawn by the end of the 12-month foow-up, a further two had died between the 12-month and the 3-year foow-up and one was excuded by their practice for heath reasons. Therefore, we approached 988 participants and 681 provided adequate acceerometry data ( 1 day with 540 minutes wear time) for anaysis, giving a 3-year foow-up rate of 69% (681/988). However, in reation to the initia tria participants providing 3-year outcome data, the 3-year foow-up rate was 681 out of 1023 (67%). The CONSORT fow diagram with 3-year foow-up data is shown in Figure 15, by randomised group. Data competeness Tabe 20 shows that 92% of participants (625/681) overa provided 5 days of acceerometry data at 3 years (88% of contro participants, 94% of posta deivery participants and 93% of nurse-support group participants). Objective physica activity findings Tabe 20 shows the summary measures for a three groups at each time point and Tabe 21 shows the estimates of effect for the different groups. For the main tria outcome of steps per day, both intervention groups were sti doing more than the origina tria contro group: 627 steps (95% CI 198 to 1056 steps) in the posta deivery group and 670 steps (95% CI 237 to 1102 steps) in the nurse-support group. The nurse-support and posta deivery groups combined did 648 steps per day (95% CI 272 to 1024 steps). The pattern was simiar for the tota weeky MVPA in bouts (minutes per week): 28 minutes (95% CI 7 to 49 minutes) in the posta deivery group and 24 minutes (95% CI 3 to 45 minutes) in the nurse-support group. The nurse-support and posta deivery groups combined did 26 minutes (95% CI 8 to 44 minutes). There was no difference between the groups at 3 years for sedentary time or daiy wear time. Missing data anayses Imputation anayses (see Tabe 22) presents the resuts for missing at random, using imputations based on the different assumptions detaied in the methods section. The imputation anayses show that making adjustments for missing vaues has ony a sma effect on the primary outcome, that is, the step count estimate, and does not change the interpretation. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 93

126 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS Patients randomised (n = 1023; 922 househods) Aocation Contro group participants (n = 338; 305 househods) Posta intervention patients (n = 339; 307 househods) Nurse intervention patients (n = 346; 310 househods) 3-month foow-up and anaysis Participants anaysed (287 househods) with compete acceerometer data (n = 318) Withdrawn, n = 1 Not abe to be contacted, n = 2 Inadequate acceerometry, n = 17 Participants anaysed (289 househods) with compete acceerometer data (n = 317) Withdrawn, n = 3 Not abe to be contacted, n = 1 Inadequate acceerometry, n = 18 Participants anaysed (286 househods) with compete acceerometer data (n = 319) Withdrawn, n = 8 Not abe to be contacted, n = 3 Inadequate acceerometry, n = month foow-up and anaysis Participants anaysed (292 househods) with compete acceerometer data (n = 323) Withdrawn, n = 3 Not abe to be contacted, n = 3 Inadequate acceerometry, n = 9 Participants anaysed (283 househods) with compete acceerometer data (n = 312) Withdrawn, n = 12 Not abe to be contacted, n = 4 Inadequate acceerometry, n = 11 Participants anaysed (289 househods) with compete acceerometer data (n = 321) Withdrawn, n = 17 Not abe to be contacted, n = 1 Inadequate acceerometry, n = 7 3-year foow-up and anaysis Participants anaysed (196 househods) with compete acceerometer data (n = 214) Withdrawn, n = 11 Not abe to be contacted, n = 10 No foow-up data provided, n = 97 Inadequate acceerometry, n = 6 Participants anaysed (216 househods) with compete acceerometer data (n = 236) Withdrawn, n = 17 Withdrawn by GP, n = 1 Not abe to be contacted, n = 9 No foow-up data provided, n = 75 Inadequate acceerometry, n = 1 Participants anaysed (211 househods) with compete acceerometer data (n = 231) Died, n = 2 Withdrawn, n = 20 Not abe to be contacted, n = 13 No foow-up data provided, n = 78 Inadequate acceerometry, n = 2 FIGURE 15 The PACE-UP CONSORT fow diagram with 3-year foow-up data. 94 NIHR Journas Library

127 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 20 Summary means and SDs for acceerometry data at baseine, 3 months, 12 months and 3 years Randomised group Contro Posta deivery Nurse-support Acceerometry data Baseine 3 months 12 months 3 years Baseine 3 months 12 months 3 years Baseine 3 months 12 months 3 years Number of participants Number of participants with 5 days wear (%) 338 (100) 286 (90) 300 (93) 188 (88) 339 (100) 282 (89) 287 (92) 222 (94) 346 (100) 296 (93) 302 (94) 215 (93) Daiy step count, mean (SD) 7379 (2696) 7327 (2688) 7246 (2671) 7281 (2721) 7402 (2476) 8086 (3014) 8010 (2922) 7896 (2853) 7653 (2826) 8707 (3206) 8131 (3228) 8131 (3410) Tota weeky minutes of MVPA in 10-minute bouts, mean (SD) 84 (97) 87 (101) 89 (94) 94 (102) 92 (90) 136 (125) 129 (124) 132 (124) 105 (116) 164 (154) 138 (141) 138 (161) Daiy sedentary time (minutes), mean (SD) 613 (68) 614 (70) 616 (72) 615(71) 614 (71) 614 (74) 617 (71) 617 (75) 619 (78) 613 (77) 620 (79) 620 (69) Daiy wear time (minutes), mean (SD) 789 (73) 795 (78) 791 (76) 789 (78) 787 (78) 798 (84) 800 (80) 798 (86) 797 (84) 805 (85) 807 (89) 805 (81) Note Acceerometry data are adjusted for day of the week and day order of wearing the acceerometer as fixed effects and participant as a random effect in a mutieve mode. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 95

128 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS TABLE 21 Acceerometry outcome data at 3 months, 12 months and 3 years. Anaysed using a avaiabe data at each foow-up. N = 954 at 3 months, N = 956 at 12 months and N = 681 at 3 years Comparison of change from baseine between randomised groups Posta deivery vs. contro Nurse support vs. contro Nurse support and posta deivery vs. contro Outcome Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue Daiy step count 3 months 692 (363 to 1020) < (844 to 1501) < months 642 (329 to 955) < (365 to 989) < (389 to 930) < years 627 (198 to 1056) (237 to 1102) (272 to 1024) < Tota weeky minutes of MVPA in 10-minute bouts 3 months 43 (26 to 60) < (44 to 78) < months 33 (17 to 49) < (19 to 51) < (20 to 48) < years 28 (7 to 49) (3 to 45) (8 to 44) Daiy sedentary time (minutes) 3 months 2 ( 12 to 7) ( 16 to 3) months 1 ( 8 to 10) ( 9 to 9) ( 7 to 8) years 1 ( 12 to 11) ( 14 to 9) ( 11 to 8) 0.77 Daiy wear time (minutes) 3 months 2 ( 8 to 12) ( 6 to 14) months 9 ( 1 to 19) ( 1 to 19) (0 to 18) years 8 ( 5 to 20) ( 6 to 19) ( 4 to 18) 0.21 Notes A modes incude treatment group, practice, sex, age at randomisation and month of baseine acceerometry as fixed effects and househod as a random effect in a mutieve mode. The xtmixed command in Stata version 12 was used, foowed by the postestimation command pwcompare to generate the pairwise estimates of effect and their CIs (see Tabe 22). The missing-not-at-random anayses make a bigger impact, but ony when we assume that there is a strong differentia departure between the non-random effects in the contro and treatment groups (soid ines in Figure 16). Even then, it is ony when we assume that the missing data in the treatment groups are 1000 steps beow their missing-not-at-random vaues, whie the vaues in the contro group are at their missing-notat-random vaues, that the treatment effects become non-significant; even then, the CI is sti argey positive. Resuts for the quaitative evauation Between March and Apri 2016, 105 participants were randomy seected, 96 were contacted and a agreed to participate. Teephone interviews were arranged and undertaken with 60 participants (20 from each tria arm). Fifty-two participants were white and eight were non-white. Interviews asted between 4 and 22 minutes (median 10 minutes). One participant who had difficuty hearing was e-maied the questions to compete. In the quotations that foow, ID3Y_ refers to the participant s ID number, F/M refers to the participant s sex, the number foowing this refers to their age and N, P or C refers to whether the participant is in the nurse-support group, the posta deivery group or the contro group, respectivey. 96 NIHR Journas Library

129 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 TABLE 22 Imputation anayses for the 3-year acceerometry outcomes Randomised group(s) vs. randomised group Posta deivery vs. contro Nurse support vs. contro Nurse support and posta deivery vs. contro Missing at random imputation anayses N Effect (95% CI) p-vaue Effect (95% CI) p-vaue Effect (95% CI) p-vaue A participants with foow-up data Imputed using treatment group, baseine steps, sex, age, practice, month of baseine acceerometry Imputed using treatment group, baseine steps, sex, age, practice, month baseine acceerometry, NS-SEC, baseine sef-reported pain and baseine body fat mass a Imputed using treatment group, baseine steps, sex, age, practice, month baseine acceerometry and 12-month steps b (198 to 1056) (174 to 1020) (211 to 1057) (217 to 1033) (237 to 1102) (268 to 1089) (293 to 1178) (270 to 1095) (272 to 1024) (295 to 1003) (239 to 1034) (305 to 1005) < < < a Baseine data for NS-SEC or sef-reported pain or body fat mass were missing for 27 participants and imputations were not avaiabe for these 27 participants when incuding these variabes as predictors. b Tweve-month steps were missing for 58 participants and imputations were not avaiabe for these 58 participants when incuding 12-month steps as a predictors. Factors affecting physica activity eves and maintenance at 3 years A key theme that emerged from our interviews was the impact that the PACE-UP tria had on participants. Most participants, regardess of what group they were in, reported an increased awareness of PA. Participants described an increased understanding about the importance of PA for heath, as we as an awareness about the amount of PA required to meet their daiy step count target: It s made me more aware of the need to actuay commit to doing some exercise a day, just stroing around the house, and going to the shops occasionay doesn t reay make much difference. It doesn t meet the sort of threshod that you need to reach to ensure that you ead a heathy ifestye. ID3Y27M56P Participants fet that taking part in the PACE-UP tria and using a pedometer had kick-started reguar activity: It was the PACE-UP tria that heped get me started and I think that did make a huge difference to me. ID3Y47F51N Participants highighted different barriers to, and faciitators of, being abe to stay physicay active in the onger term. These barriers and faciitators were often the opposite of each other; for exampe, some participants saw good weather as a motivator to engage in PA, whereas others saw bad weather as a barrier to being physicay active. Other important faciitators and barriers incuded heath, sef-motivation, ageing and socia support. These factors were considered to be important by participants, regardess of which group they were in. For some participants, they fet that engaging in reguar PA heped them to manage their heath condition: The more active I am, the better the arthritis is. ID3Y25F60P Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 97

130 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS (a) 1000 Treatment effect (95% CI) Posta deivery group ony Contro group ony Posta deivery and contro groups (b) Assumed difference from missing-at-random estimates Treatment effect (95% CI) Nurse-support group ony Contro group ony Nurse-support and contro groups (c) Assumed difference from missing-at-random estimates 1000 Treatment effect (95% CI) Posta deivery and nurse-support groups Contro group ony Posta deivery group pus nurse-support group and contro group Assumed difference from missing-at-random estimates FIGURE 16 Sensitivity anayses for different vaues of missing step counts. (a) Posta deivery group vs. contro group; (b) nurse-support group vs. contro group; and (c) posta deivery and nurse-support groups combined vs. contro group. The figures show how different vaues for the missing step counts changes the treatment effects. The starting point where a missing step counts are repaced by missing-at-random estimates in each group is represented by the zero difference estimate. Estimates in the different groups are then atered differentiay over a range of scenarios. Contro group missing step counts are the same as the 3-year missing-at-random estimates. Treatment groups missing step counts are 500 or 1000 steps ower than the 3-year missing-at-random estimates. Contro group missing step counts are 500 or 1000 steps higher than the 3-year missing-at-random estimates. Treatment group missing step counts are the same as the 3-year missing-at-random estimates. Missing step counts are 500 or 1000 steps ower or higher than the 3-year missing-at-random estimates in a contro and treatment groups (see White 191 for methods). The vertica ines are the 95% CIs for the estimated treatment effects. The treatment effect becomes statisticay not significant when the CI crosses the horizonta ine at zero. 98 NIHR Journas Library

131 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Conversey, others fet that having a heath condition was a contraindication to PA: I ve got an ongoing probem where I get pain, so there s no way I m going to be going out waking if I don t have to. ID3Y24F61P Sef-motivation was seen to be an important determinant to PA: I think it s got to come from inside. ID3Y19F62C Some participants attributed the PACE-UP tria to providing them with this motivation: Before the PACE-UP tria, I had no incentive. And that reay did hep me. That put me/gave me the first steps as it were, got me on the right track. ID3Y47F51N On the other hand, a few participants fet that since the PACE-UP tria had ended, their PA eves had decreased as a resut of a ack of sef-motivation: I m not covering anything ike what I was covering when I was on the programme, but I don t know how... to put mysef in that mind set... ID3Y55M68N Some participants chose to engage in PA as a way to stay young and sow down the ageing process: I ve got nieces and a nephew... I need to be active to keep up with them because they are young. I just need to keep up with everyone ese reay. I don t want to sow down and become od. Unfortunatey, I m not reay motivated by anything ese F65N Other participants fet that their age had become a barrier to PA and were ess ikey to do as much as they did when they were younger: If anything I m getting a bit oder and I m beginning to find it a itte bit more of a strain. ID3Y42F59N Many participants spoke about the importance of having friends and famiy to motivate them to participate in PA. Support from, and accountabiity to, famiy and friends were therefore seen as common faciitators of PA: Maybe my motivation is not ony my heath, but having somebody to do it with... to maybe be paired up with somebody who was ike-minded... if I ve promised/ony promised mysef, then I might find excuses not to do it. ID3Y50F63N Lack of socia support meant that some participants did not engage in PA: I haven t got anyone to wak with. ID3Y37F67P Lack of time was the most frequenty cited barrier to maintaining PA. Unike the other factors previousy mentioned, having time to engage in PA was not mentioned as a faciitator. The reasons for a ack of Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 99

132 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS time incuded having famiy responsibiities (ooking after chidren, caring for oder reatives), work commitments or simpy being too busy : I ve got penty of things that encourage me, it s just the time I find because I work fu time, I just find it difficut to come home, sort of prepare meas, go to the gym, go for a wak... I don t think I need any more actua motivation, I just need a bit more time! ID3Y43F54N Some participants spoke about strategies they adopted to overcome the barrier of ack of time, either by incorporating PA into their daiy routine or by buiding up their daiy PA in short bouts of activity: I wak up the escaators to get my itte bit of exercise. ID3Y15M62C As opposed to the mind set of, oh, I ve got to do an hour in the gym. Actuay 10 minutes soid waking somewhere, severa times a day, actuay buids stuff up. ID3Y56F68N At the end of the interview, participants were asked for their views on what additiona support coud be provided to aid PA maintenance. Participants were offered exampes to comment on, which incuded reguar text messages, onine resources, annua nurse appointments and waking groups. Participants had varying preferences over which additiona resources they woud find the most beneficia. Athough some participants iked the idea of a reguar text message or joy itte reminders (ID3Y47F51N) encouraging PA, others fet that these woud be too intrusive (ID3Y27M56P). Simiary, some participants iked the idea of having PA resources onine to open at their own time (ID3Y9F62C), whereas others fet that obtaining this information onine woud require a more proactive (ID3Y47F51N) approach. On the whoe, waking groups and nurse appointments were considered to be favourabe. Some participants fet that waking groups woud provide more motivation to go out (ID3Y17M68C) and a reguar nurse or other appointment woud provide externa accountabiity (ID3Y52M66N). As we as providing feedback on suggested possibe resources, we proposed that participants aso come up with additiona suggestions. These suggestions incuded hoistic appointments with a nurse to discuss both diet and PA, and more opportunities for oder peope: There is not much for peope over 60, there s no rea paces that are easy on the doorstep. ID3Y2F64C One participant spoke about being afraid to increase their PA eves, as they were unsure of whether or not it was safe. This participant sought more guidance around riskess ways to increase PA eves: I m doing quite we with what I m doing, so what s the point of sort of having a risk of a heart attack or something ike that, suddeny break in to a stride and start running, so maybe a bit of guidance on what s going to happen to you if you do step up your exercise pan. ID3Y23M67P The effect of the minima intervention on physica activity eves of participants in the contro group Of the 20 contro group participants interviewed, 17 received the pedometer, handbook and diary after 12 months. Thirteen of these participants reported using these resources when they first received them; however, at 3 years, ony four participants were continuing to monitor their steps [two with a pedometer, one with a Fitbit (Fitbit, San Francisco, CA, USA) and one with a mobie phone]. 100 NIHR Journas Library

133 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Athough most of the contro group participants who were interviewed fet that the PACE-UP tria had not increased their PA eves, they sti stated that the PACE-UP tria had increased their awareness: It s made me aware that I do not do as much as I shoud be doing. ID3Y20F66C Some, however, did tak about changes they had made to their daiy ives as a resut of their increased awareness: I m more ikey to wak to work now, rather than going on the bus. ID3Y11M54C For those who did not utiise the pedometer, some reported difficuties in using it: I coud not work the pedometer... coud not get it going. ID3Y9F63C A few others who used the pedometer stated that discontinued use was either because it had a negative impact on them psychoogicay or because they had faen out of the habit of using it reguary: It s quite distressing to see how itte I do.... there are ots of other things [that] intrude and you tend to sip back to od patterns. ID3Y1M58C ID3Y5F68C There was aso a suggestion that, as we as the pedometer, you needed to have someone to report back to: It aways needs to have someone keeping you aware I think. ID3Y5F68C Discussion Main quantitative findings from the 3-year foow-up We foowed up just over two-thirds of the origina tria cohort with acceerometry outcome data at 3 years (and over 90% provided 5 days of data). Compared with baseine, those in the origina nurse-support and posta deivery groups were sti doing significanty more steps per day and weeky time in MVPA in bouts at 3 years than the contro group. There were no significant differences in outcomes between the posta deivery and nurse-supported intervention groups (as was aso the case at 12 months), and there were no significant differences between the three groups in terms of wear time or sedentary time. Our sensitivity anayses ooking at the potentia impact of missing outcome data at 3 years suggest that it is highy unikey that missing data have substantiay biased our resuts. Fairy extreme departures from the missing-at-random anayses were needed to resut in non-significant effects and, even then, the 95% confidence imits were argey positive. This suggests that the tria interventions had a persistent effect on objectivey measured PA eves at 3 years, with no difference between intervention groups. The fact that the minima intervention given to the contro group at 12 months was not effective at increasing the participants PA eves suggests that the additiona support given to the origina tria posta intervention group (with a foow-up teephone ca after 1 week and encouragement to return the competed PA diary after 3 months) was an important component of this group s success. Both the foow-up teephone ca and the encouragement to return the competed PA diaries after the intervention were not part of the intended intervention package, but, rather, were research measures as part of the process evauation to Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 101

134 THREE-YEAR FOLLOW-UP TO ASSESS THE MAINTENANCE OF PHYSICAL ACTIVITY LEVELS ensure fideity of the intervention deivery. However, this minima support, which was not provided face to face or by a heath-care professiona, seems to have been important to the success of the posta intervention. The origina tria posta deivery group aso received the posta pedometer intervention when they had just been recruited to the PA tria, when motivation may have been higher, and they had step count targets set for them based on their baseine binded pedometer use, whereas those receiving the materias at 12 months needed to wear the pedometer again for 1 week to set their target step count. These factors may aso have been important to the success of the tria posta intervention group. Main quaitative findings from the 3-year foow-up interviews A key finding was that most participants discussed their increased awareness of PA, irrespective of which group they were in and regardess of whether or not they thought that the PACE-UP tria had actuay increased their PA eves. Key barriers to, and faciitators of, maintaining PA were reported that were often the inverse of one another and incuded heath, weather, sef-motivation, ageing and socia support. Lack of time was the most frequenty cited barrier. Some participants were abe to overcome ack of time by incorporating PA into their daiy routine or by breaking PA down into smaer, more manageabe bouts throughout the day. Participants gave us mixed feedback on how usefu they thought text messages and onine resources woud be to hep to inform future interventions to increase and maintain PA, but waking groups and nurse or other appointments to provide externa accountabiity were broady wecomed. Additiona suggestions provided incuded more hoistic appointments with a nurse and more opportunities for PA for oder peope. Participants had differing opinions over the resources they woud find most beneficia, emphasising the importance of individua taioring of some aspects of PA interventions. Ony a few of the contro group participants interviewed were continuing to use the pedometer provided at 12 months. Some participants were not sure how it worked, and others fet that, as we as using the pedometer, it was important to have someone to report back to. This highights the importance of the extra contact that participants in the posta deivery group received as part of the main tria: a foow-up teephone ca to check that they knew how to work the device and encouragement to send back their competed PA diaries with step count recordings after the 12-week programme. Further discussion of the strengths and weaknesses of both the quantitative and quaitative approaches, and the impications of the findings for heath care and future research, are detaied in Chapter NIHR Journas Library

135 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Chapter 9 Discussion Summary of the findings The PACE-UP tria demonstrated that both the posta deivery and the nurse-supported pedometer interventions, based on trying to graduay add in 3000 steps in 30 minutes on most days, increased objectivey assessed PA (step counts by around one-tenth and MVPA in bouts by around one-third) among predominanty inactive 45- to 75-year-ods at 12 months. Athough the nurse-ed deivery had a greater effect than the posta deivery at 3 months, by 12 months this difference was not sustained. The interventions had no effect on sedentary time, anthropometry or other outcomes and did not increase AEs. No effect modification was demonstrated (by age, sex, taking part as a coupe, sef-efficacy, disabiity, socioeconomic status, pain or BMI). Questionnaire-based outcome measures tended to support the concusions of acceerometer measures, but ony if waking was an expicit part of the questionnaire. Thus, the IPAQ MVPA question did not show any intervention effect, but the IPAQ waking question showed a significant effect of both the nurse-supported and the posta deivery interventions with no difference between the interventions, athough with ess precision than the acceerometry data. Both interventions were we accepted and the tria had high fideity; three-quarters of the nurse group attended a three sessions and around 80% of both the posta deivery and the nurse-support groups returned competed step count diaries. Increase in step count was positivey associated with both nurse session attendance and competed diary return. Incrementa cost per step was 0.19 and 3.61 per minute in a 10-minute MVPA bout for the nurse-support group, whereas the posta deivery group took more steps and cost ess than the contro group. The posta deivery group had a 50% chance of being cost-effective at a 20,000 per QALY threshod within 1 year. The QALY-based concusion changed to the contro group dominating the posta deivery group when four aternative assumptions were made (using the 3-month outcome data, extending the perspective to participants, excuding heath service use and using sef-reports of AEs), athough this was not the case for cost-effectiveness ratios based on step count and MVPA. The posta deivery group was significanty more cost-effective than both the nurse-support group and the contro group in the ong term and this finding was robust to changes in assumptions. Nurses and intervention group participants described the intervention in a positive way and confirmed that primary care was an appropriate setting. Nurses beieved that participating in the tria, especiay in the BCT training, enhanced the quaity and deivery of the advice and support they provided within routine consutations. Participants described important faciitators of increasing PA, incuding the desire for a heathy ifestye, improved physica heath, enjoyment of waking in the oca environment, having a fexibe routine, appropriate sef-monitoring and externa monitoring and support from others. Important barriers incuded physica heath probems, having an infexibe routine, work and other commitments and poor weather. Severa BCTs were highighted as having an important impact, incuding sef-monitoring and review of goas and outcomes, panning socia support/change and reapse prevention. Athough most participants in the posta deivery group were confident in increasing their PA without nurse support, two key caveats were existing heath probems and overcoming barriers. The foow-up of over two-thirds of the tria cohort at 3 years demonstrated persistent increases in both step count and time in MVPA for the nurse-support and posta deivery groups compared with the contro group, with no difference between intervention groups. The posta intervention given to the contro group at 12 months, with no foow-up teephone ca after 1 week and no requirement to post back the diary to be reviewed after 3 months, was not effective at increasing participants PA eves. This suggests that these minima support components of the posta deivery intervention, which were not face to face or not Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 103

136 DISCUSSION provided by a heath-care professiona, may have been important to its success. Quaitative evauation found that most participants fet that the PACE-UP tria had increased their awareness of the importance of PA, irrespective of their intervention group and whether or not they fet that their PA had actuay increased. Many of the barriers to, and faciitators of, PA maintenance were the inverse of each other and most were simiar to those found to be important for increasing PA during the actua tria (heath conditions, weather, ageing, socia support, time). Participants varied in the resources that they woud find most beneficia to hep them maintain their PA eves, emphasising the importance of individua taioring of some aspects of PA interventions. The PACE-UP tria was nove in ceary separating out the effects of pedometer provision and nurse support in a genera popuation sampe of aduts and oder aduts and in demonstrating the effects on both step counts and MVPA in bouts, thus making the outcome assessment reevant to current nationa and internationa PA guideines. Strengths and imitations Study strengths The PACE-UP study had many important strengths. It was arge and popuation based with primary care samping, aowing response and any bias in response to be assessed, rather than reying on recruiting vounteers. It was designed to have househod randomisation, which aowed two members of a coupe to take part together if they wished to, enabing a comparison of individua and coupe effects. It had three tria arms, aowing the separation of nurse support and pedometer/handbook/diary effects. The intervention was pragmatic, using practice nurses who worked in the practices to deiver the nurse PA consutations, rather than externa researchers or exercise speciaists. There was a very good uptake rate of nurse appointments and return of competed step count diaries, showing participant engagement with the interventions. The main PA outcomes were objectivey measured and were reevant to the PA guideines. AEs were measured in a number of ways to minimise bias, both sef-reported from questionnaires and objectivey from primary care records. The tria achieved a foow-up rate of over 90% with compete primary outcome data. The tria aso incuded embedded process, quaitative and economic evauations, with the economic evauations using tria resuts in a simuation of ong-term cost-effectiveness. An extended 3-year foow-up aowed maintenance of any intervention effects beyond 12 months to be studied. Study imitations There were aso some important study imitations. The 10% recruitment into the tria is considered in detai beow, in Generaisabiity. At the baseine assessment, 218 out of 1023 participants (21%) achieved the PA guideine targets based on their acceerometry. These participants were not excuded from the tria because if the intervention were to be roed out in primary care, sef-reported PA eves woud define participation. Our nurse-supported intervention group had sighty higher baseine PA eves; however, the tria resuts were not biased, as anayses were based on individua change, controing for baseine PA eve. It was impossibe to mask participants and nurses to the intervention group and, pragmaticay, research assistants recruited and foowed up the same participants, so were unmasked to group at the outcome assessment. However, a the primary and secondary PA outcomes were assessed objectivey by acceerometry. It is possibe that participants might have tried harder with their PA when monitored, but this woud aso have affected contro participants and woud be reduced by using a 7-day protoco for data coection. 31 In addition, our intervention groups increased their MVPA in bouts of 10 minutes, impying that participants made changes suggested by the programme. Despite recruiting to target and having exceent foow-up, our CIs for the difference between intervention groups cannot rue out a sma 12-month difference. Interpretation of our 3-year foow-up findings was potentiay imited by the fact that two-thirds of the contro group participants received a pedometer, handbook and diary, and 20% of them aso received a singe nurse appointment after their 12-month foow-up. However, any contamination appears to be minima, as there was no evidence of a change in the contro group and the intervention estimates at 3 years were very simiar to those at 12 months. These findings are of potentia importance as, in combination, they suggest that the minima contact with the 104 NIHR Journas Library

137 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 posta deivery group after participants were sent their pedometer packs (teephone contact after 1 week and encouragement to return their competed step count diaries at 3 months) was important in stimuating an effect. Timing may have been important too; as the contro group was offered the intervention 12 months after the participants had initiay expressed an interest in participating in the tria, they woud have had to have worn the pedometer to measure their step count and set targets, whereas the tria posta deivery group had targets set based on wearing a binded pedometer at baseine. These added factors may have aso been important to the success of the tria posta deivery intervention group. Generaisabiity Overa, ony 10% of peope invited to participate in the tria ended up being recruited and randomised. This is simiar to other primary care PA trias, 33,192 but ower than the 30% that we achieved in our recent oder adut PA tria. 21 However, 10% of a popuation sampe is sti a usefu percentage to participate in a pubic heath intervention, and this tria shows the potentia of primary care to contribute to PA pubic heath goas, particuary within an urban context. As we as monitoring overa recruitment, using primary care as a samping framework aowed us to ook for any seection bias in recruitment to the tria. Primary care record comparisons showed that participation rates were significanty ower in men, in those aged < 55 years, in those who were iving in the most socioeconomicay deprived quintie and among Asian rather than white or back ethnic groups. Despite seecting practices from deprived, ethnicay diverse areas, few participants were from ower socioeconomic and ethnic minority groups, imiting both the subgroup anaysis power and the generaisabiity to more diverse popuations. Faiure to incude socioeconomicay deprived or ethnic minority groups, in which PA eves were ower, coud aso increase heath inequaities. In a RCT of an intervention, it is not possibe to separate out reuctance to participate in the intervention from reuctance to participate in the tria itsef, with requirements for informed consent, randomisation and rigorous foow-up and evauation. If the intervention were to be roed out in routine primary care, uptake coud be higher and ess prone to seection bias. Handing out the intervention materias (pedometer, handbook and diary) in primary care consutations in which advice to increase ow PA eves is aready being offered may aso increase the intervention s reach (e.g. in reevant chronic disease consutations, or as part of NHS Heath Checks, which cover a simiar age group and aim to reduce cardiovascuar risk 59 ). The intervention coud aso be a vauabe addition to diabetes meitus prevention strategies, such as the NHS Diabetes Prevention Programme, whereby primary care is being used to identify patients at a high risk of deveoping diabetes meitus, many of whom are inactive 193 and with higher proportions from ethnic minority groups. Using the PACE-UP tria intervention in these ways woud need further evauation and monitoring, but this may have the potentia to improve generaisabiity and either decrease, or at east not increase, inequaities. Comparison with other studies We beieve that this is the argest popuation-based tria of a pedometer-based waking intervention with 12-month foow-up findings and the ony pedometer tria with objective PA data on time in MVPA, which is reevant to PA guideines at 3 years. The resuts are consistent with, and extend, our findings in 60- to 75-year-ods that were achieved in the smaer PACE-LIFT tria 21 and aso support the recent change in NICE guidance to promoting pedometers as part of packages incuding support to set reaistic goas, monitoring and feedback. 37 The intervention used in the PACE-LIFT tria aso incuded pedometer feedback, use of a step count diary and practice nurse PA consutations based around BCTs. However, the PACE-LIFT tria intervention comprised four onger practice nurse consutations, which aso incuded individua acceerometer feedback on PA intensity. The PACE-LIFT tria was a two-arm tria with ony a singe intervention arm, and was therefore unabe to separate out PA monitor effects from those of the nurse-support group. Despite incuding a much ess intense intervention, the PACE-UP tria has deivered simiar eves of effect at both 3 and 12 months in PA outcomes and, furthermore, has shown what can be achieved via a posta route. It is aso reassuring that our interventions did not increase sedentary time, given its potentia harm, 194,195 as compensation can sometimes occur. The absoute step count increase Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 105

138 DISCUSSION achieved in the PACE-UP tria was modest compared with that reported in systematic reviews. 26,31,32 However, most trias with 12-month data have been based on sma numbers and recruited either vounteers 196 or high-risk groups, 35 or reported ony sef-reported PA data; 97 a of these factors are ikey to ead to arger effect sizes. Athough PA guideines focus on time in MVPA in bouts, not on step counts, the systematic reviews presented no data on this important outcome. 26,31,32 The PACE-UP tria resuts confirm the PACE-LIFT tria findings, 21 with significant 12-month increases in MVPA in bouts. Based on the 3000 in 30 formua, extra minutes of MVPA per week in bouts corresponds to approximatey 500 extra steps per day. Thus, approximatey three-quarters of the extra steps achieved in the PACE-UP tria ( per day) contributed to an increase in MVPA in bouts. We beieve that our tria is the first to show that the 3000-in-30 message 43 can ead to an approximatey one-third increase in weeky MVPA in bouts at 12 months, as was achieved across both intervention groups. The 3000 steps in 30 minutes formua neaty captures intensity 43 and coud become an important new pubic heath goa, particuary as many peope now have the abiity to measure steps easiy with their mobie phones. Based on a systematic review, which has quantified the strength of association between waking and the risk of deveoping CHD, 197 the increase of 33 minutes per week in the posta deivery group in our study at 12 months, if sustained, woud be expected to reduce the risk of CHD by approximatey 4.5% (95% CI 3% to 6%; see Appendix 8 for detais). A cohort study that has reated pedometer-measured steps to mortaity 198 has simiary aowed us to estimate that a sustained increase of 642 steps per day woud be expected to ead to a decrease in a-cause mortaity of approximatey 4% (95% CI 1% to 7%; see Appendix 8 for detais). Recacuating these estimates for the effect estimates in the posta deivery group at 3 years makes itte difference, with the resutant decreases being 4% (95% CI 3% to 5%) for CHD and 4% (95% CI 1% to 5%) for tota mortaity. Most pedometer-based interventions have not separated out the effects of the pedometer itsef from the effects of the additiona support provided. 21,24,31 The Heathy Steps tria 97 showed that pedometers achieved an additiona effect compared with a primary care green prescription, but the PA outcomes presented were based on sef-report. The PACE-UP tria demonstrates that athough the nurse intervention group had a significanty greater effect on both step counts and time in MVPA at 3 months, by 12 months both the nurse-supported and posta deivery interventions sti had a significant effect, but with no evidence of a difference between them. This stronger effect during the period of contact with the nurse, which was not sustainabe in the onger term, has aso been shown in other interventions with heath professionas. 199 Both the nurse-support and posta deivery groups received a pedometer, diary and handbook as part of the PACE-UP tria package; therefore, it is not possibe to know how much the individua components contributed. A systematic review suggested that step count diaries were common to successfu pedometer interventions, 31 and approximatey 80% of both of our intervention groups returned competed step count diaries. In addition, our process evauation showed that returning a competed diary was significanty associated with an increase in step counts for both of the intervention groups. Quaitative findings aso confirmed that participants from both groups vaued the handbook and diary, as we as the pedometer. 162 Contro group participants provided with the pedometer, diary and handbook by post at 12 months did not significanty increase their PA eves; however, they were not asked to return competed step count diaries after 3 months, which may have contributed to the ack of effect of the materias in this group. We found no effect of the interventions on anthropometric measures, such as BMI or fat mass; this is consistent with other simiar studies. 21,196 Our interventions aso did not affect anxiety or depression scores, which is consistent with other primary care pedometer-based interventions, suggesting either no effect or insensitivity of these measures to change, particuary when eves are in the norma range for most peope. 21,33 Athough a few participants mentioned that they had negative effects from overdoing waking, most intervention participants taked about feeing fitter, seeping better, improved mood, having more energy, ess pain and keeping more active into oder age. 162 There is currenty a ack of data comparing individua, coupe or househod participation in waking studies. 21,30 Househod samping aowed us to investigate this in the PACE-UP tria, but, unfortunatey, ony 20% of the participants took part in the study as a coupe; therefore, we had reduced power for our subgroup anaysis, which showed no effect of taking part as a coupe, simiar to the findings in our PACE-LIFT tria NIHR Journas Library

139 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 The sef-efficacy differences that we demonstrated between both intervention groups and the contro group at 3 months and between the nurse-support group and contro group participants at 12 months are consistent with the positive reationship between changing sef-efficacy and PA behaviour that others have reported. 200 The BCTs most associated with sef-efficacy and successfu PA outcomes are goa- and action-panning, prompting sef-monitoring and feedback and panning of socia support/change. 200 A of these BCTs were specificay recommended in recent guidance 29 and were incuded in our study in written materias for both intervention groups and as a focus of nurse PA consutations. 70 Our quaitative interviews found that more BCT comments were made by the nurse support group than by the posta deivery group, apart from around sef-monitoring. 162 Increased sef-efficacy has aso been shown to be important for ong-term PA adherence; 201 however, we found no difference in 3-year PA maintenance between intervention groups, despite the nurse-support group having a higher eve of sef-efficacy than the posta group at 12 months. Waking is a safe intervention, which is indicated in many chronic diseases, 1,8 athough empirica data on the safety of waking interventions are imited. 24 A arge tria based on 40- to 74-year-od women, which encouraged a singe 30-minute brisk wak 5 days weeky, reported increased fas and injuries. 60 Our findings in the PACE-UP tria, showing no increase in AEs, buids on simiar evidence from the PACE-LIFT tria, 21 using both sef-reported and objective primary care data, and highights the potentia importance of buiding up MVPA graduay, particuary in oder aduts, those who are inactive or those who have comorbidities. 1,11 The suggestion of a protective effect of the interventions in the PACE-UP tria on both fas and cardiovascuar events at 12 months is pausibe, but not definitive, as it is based on ony a sma number of events. We demonstrated a persistent intervention effect at 3 years, in terms of both step counts and time in MVPA in bouts. This adds to the imited evidence from the systematic review by Hobbs et a., 26 who found ony two trias with objective PA measurement data beyond 12 months in this age group. One tria reported a significant intervention effect on step counts at 18 months, but suffered high attrition bias, 202 and another tria found no effect at the 24-month foow-up on either step count or acceerometry-assessed vector magnitude. 203 Recent NICE guidance 29 and a Cochrane systematic review 23 aso caed for PA interventions with a onger foow-up period and objective PA measures. Our quaitative evauations at 3 years aso add to the imited evidence base on factors that ead to successfu and unsuccessfu PA maintenance 190 and suggest that many of the factors that were important barriers to, or faciitators of, increasing PA eves in the origina tria (e.g. heath conditions, weather, ageing, socia support, time) are sti important when considering maintenance. This provides support for the credibiity of our work and suggests that barriers and faciitators may be simiar for both PA adoption and maintenance. 204 Our findings support others in suggesting that future interventions shoud focus on techniques to transform PA barriers into faciitators, for exampe by demonstrating the vaue of PA for many chronic heath conditions, as we as safe ways in which individuas can increase their PA eves to change the presence of chronic heath conditions from inhibiting to promoting PA as peope age. 157,205,206 Our resuts on cost-effectiveness provide new evidence in a research area that refects a dearth of primary evidence. 139,140 The evidence that the posta intervention has a 50% chance, and the nurse intervention a 5% chance, of being cost-effective within 12 months is new. Athough ower than the 95% ikeihood of green prescriptions being cost-effective in New Zeaand at 12 months, 97 it is sti a reasonaby high percentage for a behaviour change intervention to achieve at 12 months. The expectation that the posta intervention is most cost-effective over a ifetime is very strong and is comparabe to other findings from modes. 141,142 Interpretation of the resuts Primary care patients aged years can achieve important increases in their PA eves using a 12-week pedometer-based waking intervention, incuding handbooks and PA diaries (avaiabe at Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 107

140 DISCUSSION deivered either by post with minima support or through practice nurse PA consutations, with both methods achieving simiar 12-month effects. An important part of the intervention was to try and graduay add in 3000 steps in 30 minutes most days weeky. The persistent effect at 3 years suggests a ong-term beneficia effect. This is a safe intervention that is acceptabe to patients and nurses. The posta deivery group was significanty more cost-effective than the nurse-support and contro groups in the ong term, thus providing a cost-effective way of deivering ong-term quaity-of-ife benefits. The ack of an increase in PA eves at 3 years in the contro group, which received a simpe posta intervention without further contact after the 12-month foow-up, suggests that contacting participants after posting the intervention components to them and encouraging the return of PA diaries may be important factors for success for the posta intervention route, but this minima support does not need to be face to face or provided by a heath-care professiona. Concusions Impications for heath care A primary care pedometer-based waking intervention, deivered by post with minima support, coud provide an effective and cost-effective approach to addressing the pubic heath physica inactivity chaenge. The 3000 steps in 30 minutes formua neaty captures intensity and coud become a usefu new pubic heath goa, particuary as many peope can measure steps easiy with their mobie phones. The PACE-UP 12-week pedometer-based waking intervention coud be considered for incusion into the NHS Heath Check programme, aimed at a simiar age group (40- to 74-year-ods), and/or the NHS Diabetes Prevention Programme. Recommendations for research The PACE-UP tria generaisabiity is imited by the 10% overa recruitment rate and a ower recruitment rate in Asian and socioeconomicay deprived patients. Further research into different recruitment methods is needed, as is research assessing the recruitment rate achievabe if this programme were offered outside a tria setting over a more proonged time period. Athough the overa posta intervention outcomes were as effective and more cost-effective than the nurse-supported intervention outcomes, further research is required to understand who woud benefit most from the individua taioring offered by a nurse-supported intervention. There has been a recent dramatic increase in the use of wearabes to monitor persona PA eves, incuding through smartphones, wrist-worn devices, onine monitoring and mobie apps. Further research into how the PACE-UP 12-week PA programme coud be integrated into the use of these devices (with/without a pedometer) is needed. 108 NIHR Journas Library

141 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Acknowedgements A the authors woud ike to pay tribute to Dr Suni Shah, who was one of the tria investigators and heped to conceive the idea for the tria, and was invoved in its panning, deivery and evauation unti his death in September We are gratefu to the south-west London genera practices, their practice nurses who supported this study and a of the patients from these practices who participated: Upper Tooting Road Practice, Tooting; Chatfied Practice, Battersea; Wrythe Green Surgery, Carshaton; Francis Grove Surgery, Wimbedon; Putneymead Group Medica Practice, Putney; Heathfied Practice, Putney; and Cricket Green Medica Practice, Mitcham. We woud ike to thank our supportive TSC: Professor Sarah Lewis (chairperson), Professor Pau Litte (GP representative) and Mr Bob Laventure (patient and pubic invovement representative). We woud aso ike to provide our gratefu thanks to the foowing peope for their input in the study: Dr Iain Carey, who heped with processing the downoaded genera practice data; Debbie Brewin, who heped with practice nurse training and handbook and diary content; and Becca Homes, who conducted the non-participant interviews. Contributions of authors Tess Harris (GP and Professor of Primary Care Research) conceived the idea for the study with the other study co-investigators; ed the protoco deveopment and the funding appication; co-designed the behaviour change intervention, the patient handbook and the patient diary; supervised the running of the tria; contributed to the anaysis and ed the drafting of the report. Say Kerry (Reader in Medica Statistics) heped to conceive the idea for the study, participated in the study design and the deveopment of the research protoco and the funding appication, sat on the TMG, wrote the statistics anaysis pan, performed the quantitative anayses and contributed to the drafting of the report. Christina Victor (Professor of Gerontoogy and Heath Services Research) heped to conceive the idea for the study, deveoped the research protoco and funding appication, heped with the questionnaire deveopment, ed on the quaitative aspects of the study and contributed to the drafting of the report. Steve Iiffe (GP and Professor of Primary Care for Oder Peope) heped to conceive the idea for the study, deveoped the research protoco and funding appication, advised on the tria safety aspects, heped with the questionnaire deveopment and contributed to the drafting of the report. Michae Ussher (Professor of Behavioura Medicine) heped to conceive the idea for the study, deveoped the research protoco and funding appication, co-designed the behaviour change intervention, co-designed and heped to carry out the nurse training, advised on the questionnaire deveopment and contributed to the drafting of the report. Juia Fox-Rushby (Professor of Heath Economics) deveoped the research protoco and funding appication, designed the heath economics procedures and data coection toos, supervised the heath economics data coection, anaysis and reporting and contributed to the drafting of the report. Peter Whincup (Professor of Cardiovascuar Epidemioogy) heped to conceive the idea for the study, deveoped the research protoco and funding appication, heped with the questionnaire deveopment, advised on the anthropometric assessment and contributed to the drafting of the report. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 109

142 ACKNOWLEDGEMENTS Uf Ekeund (Professor of Physica Activity Epidemioogy) heped to conceive the idea for the study, deveoped the research protoco and funding appication, advised on PA measurement, reporting and anaysis and contributed to the drafting of the report. Chery Furness (Tria Manager) provided the day-to-day overa management of the tria, sat on the TMG, co-designed and carried out the nurse training, co-ordinated the recruitment of practices and participants, took responsibiity for data management, ed on the process evauation and contributed to the drafting of the report. Eizabeth Limb (Study Statistician) sat on the TMG, heped to write the statistics anaysis pan, arranged the random samping of househods, performed the quantitative anayses, prepared the tabes and fow diagrams and contributed to the drafting of the report. Nana Anokye (Heath Economist) designed the heath economics procedures and data coection toos, carried out the heath economics anayses, prepared the tabes and appendices reating to the heath economics resuts and contributed to the drafting of the report. Judith Ibison (GP and Senior Lecturer in Primary Care) heped with the recruitment of practices, recruitment of research staff, downoading of data in practices, questionnaire deveopment and panning of the intervention deivery and contributed to the drafting of the report. Stephen DeWide (GP and Senior Lecturer in Primary Care) heped with the recruitment of practices, advising on GP searches, downoading of data in practices, interpretation of GP data and contributed to the drafting of the report. Lee David (GP and Director of 10 Minute CBT) co-designed the behaviour change intervention, co-designed the patient handbook and diary, advised on the nurse and patient feedback forms, co-designed the nurse training, ed on the behaviour change aspects of training, provided feedback to the nurses on BCT deivery and contributed to the drafting of the report. Emma Howard (Research Assistant) was invoved in compiing patient information and data coection packs, conducted recruitment and foow-up of participants (incuding informed consent, randomisation and data coection), coated and anaysed data for the process evauation and contributed to the drafting of the report. Rebecca Dae (Research Assistant) was invoved in compiing patient information and data coection packs, conducted recruitment and foow-up of participants (incuding informed consent, randomisation and data coection) and contributed to the drafting of the report. Jaime Smith (Research Assistant) recruited and foowed up participants (incuding informed consent, randomisation and data coection), carried out quaitative interviews with the intervention group participants, undertook the quaitative anayses and contributed to the drafting of the report. Rebecca Normanse (GP and Deputy Coordinating Editor of the Cochrane Airways Group) carried out quaitative interviews with the intervention group participants and practice nurses, undertook the quaitative anayses and contributed to the drafting of the report. Caroe Beighton (Senior Lecturer in Nursing) conducted a focus group with the practice nurses, carried out quaitative interviews with participants as part of the 3-year foow-up, undertook the quaitative anayses and contributed to the drafting of the report. Katy Morgan (Statistician) carried out anayses reating to the generaisabiity and representativeness of the sampe and contributed to the drafting of the report. 110 NIHR Journas Library

143 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Charotte Wahich (Research Assistant) contacted participants for the 3-year foow-up (incuding informed consent and data coection), carried out the quaitative interviews with participants as part of the 3-year foow-up, undertook the quaitative anayses and contributed to the drafting of the report. Sabina Sanghera (Heath Economist) carried out the economic anayses reating to the heath services use data and contributed to the drafting of the report. Derek Cook (Professor of Epidemioogy) heped to conceive the idea for the study; participated in the study design, deveopment of the research protoco and the funding appication; sat on the TMG; wrote the statistica anaysis pan; performed the quantitative anayses and contributed to the drafting of the report. Pubications Harris T, Kerry SM, Victor CR, Shah SM, Iiffe S, Ussher M, et a. PACE-UP (Pedometer And Consutation Evauation UP) a pedometer-based waking intervention with and without practice nurse support in primary care patients aged years: study protoco for a randomised controed tria. Trias 2013;14:418. Normanse R, Smith J, Victor C, Cook DG, Kerry S, Iiffe S, et a. Numbers are not the whoe story: a quaitative exporation of barriers and faciitators to increased physica activity in a primary care based waking intervention. BMC Pubic Heath 2014;14:1272. Beighton C, Victor C, Normanse R, Cook D, Kerry S, Iiffe S, et a. It s not just about waking... it s the practice nurse that makes it work : a quaitative exporation of the views of practice nurses deivering compex physica activity interventions in primary care. BMC Pubic Heath 2015;15:1236. Normanse R, Homes R, Victor C, Cook DG, Kerry S, Iiffe S, et a. Exporing non-participation in primary care physica activity interventions: PACE-UP tria interview findings. BMC Trias 2016;17:178. Harris T, Kerry SM, Limb ES, Victor CR, Iiffe S, Ussher M, et a. Effect of a primary care waking intervention with and without nurse support on physica activity eves in 45- to 75-year-ods: the Pedometer And Consutation Evauation (PACE-UP) custer randomised cinica tria. PLOS Med 2017;14:e Wahich C, Beighton C, Victor C, Normanse R, Cook D, Kerry S, et a. You started something... then I continued by mysef : a quaitative study of physica activity maintenance. Prim Heath Care Res Dev 2017;18: Furness C, Howard E, Limb E, Cook DG, Kerry S, Wahich C, et a. Reating process evauation measures to compex intervention outcomes: findings from the PACE-UP primary care pedometer-based waking tria. Trias 2018;19:58. Harris T, Kerry SM, Limb ES, Furness C, Wahich C, Victor CR, et a. Physica activity eves in aduts and oder aduts 3-4 years after pedometer-based waking interventions: Long-term foow-up of participants from two randomised controed trias in UK primary care. PLOS Med 2018;15:e Kerry SM, Morgan KE, Limb E, Cook DG, Furness C, Carey I, et a. Interpreting popuation reach of a arge, successfu physica activity tria deivered through primary care. BMC Pubic Heath 2018;18:170. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 111

144 ACKNOWLEDGEMENTS Data sharing statement Anonymised individua patient data may be avaiabe for the tria effectiveness and cost-effectiveness anayses. These wi be stored in a secure data repository at St George s, University of London. A queries and requests shoud be submitted to the corresponding author for initia consideration. Patient data This work uses data provided by patients and coected by the NHS as part of their care and support. Using patient data is vita to improve heath and care for everyone. There is huge potentia to make better use of information from peope s patient records, to understand more about disease, deveop new treatments, monitor safety, and pan NHS services. Patient data shoud be kept safe and secure, to protect everyone s privacy, and it s important that there are safeguards to make sure that it is stored and used responsiby. Everyone shoud be abe to find out about how patient data are used. #datasavesives You can find out more about the background to this citation here: NIHR Journas Library

145 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 References 1. Department of Heath and Socia Care. Start Active, Stay Active: A Report on Physica Activity for Heath from the Four Home Countries Chief Medica Officers. London: Department of Heath and Socia Care; Word Heath Organization. Goba Recommendations on Physica Activity for Heath. Geneva: Word Heath Organization; Lee IM, Shiroma EJ, Lobeo F, Puska P, Bair SN, Katzmarzyk PT, Lancet Physica Activity Series Working Group. Effect of physica inactivity on major non-communicabe diseases wordwide: an anaysis of burden of disease and ife expectancy. Lancet 2012;380: /S (12) Odridge NB. Economic burden of physica inactivity: heathcare costs associated with cardiovascuar disease. Eur J Cardiovasc Prev Rehabi 2008;15: HJR.0b013e3282f19d42 5. Carson SA, Futon JE, Pratt M, Yang Z, Adams EK. Inadequate physica activity and heath care expenditures in the United States. Prog Cardiovasc Dis 2015;57: j.pcad Garber CE, Bissmer B, Deschenes MR, Frankin BA, Lamonte MJ, Lee IM, et a. Quantity and quaity of exercise for deveoping and maintaining cardiorespiratory, muscuoskeeta, and neuromotor fitness in apparenty heathy aduts: guidance for prescribing exercise. Med Sci Sports Exerc 2011;43: O Donovan G, Bazevich AJ, Boreham C, Cooper AR, Crank H, Ekeund U, et a. The ABC of Physica Activity for Heath: a consensus statement from the British Association of Sport and Exercise Sciences. J Sports Sci 2010;28: Morris JN, Hardman AE. Waking to heath. Sports Med 1997;23: NHS Digita. Heath Survey for Engand 2008: Physica Activity and Fitness. London: NHS Digita; de Souto Barreto P. Goba heath agenda on non-communicabe diseases: has WHO set a smart goa for physica activity? BMJ 2015;350:h Sparing PB, Howard BJ, Dunstan DW, Owen N. Recommendations for physica activity in oder aduts. BMJ 2015;350:h NHS Digita. Heath Survey for Engand London: NHS Digita; Nationa Centre for Socia Research. Heath Survey for Engand London: Department of Heath; Word Heath Organization. Goba Heath Observatory Data Repository. URL: ncd/risk_factors/physica_activity_text/en/ (accessed 25 May 2018). 15. Tudor-Locke CE, Myers AM. Chaenges and opportunities for measuring physica activity in sedentary aduts. Sports Med 2001;31: Morey MC, Suivan RJ. Medica assessment for heath advocacy and practica strategies for exercise initiation. Am J Prev Med 2003;25(Supp. 3): S (03) Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 113

146 REFERENCES 17. Cress ME, Buchner DM, Prohaska T, Rimmer J, Brown M, Macera C, et a. Best practices for physica activity programs and behavior counseing in oder adut popuations. J Aging Phys Act 2005;13: Thompson PD, Buchner D, Pina IL, Baady GJ, Wiiams MA, Marcus BH, et a. Exercise and physica activity in the prevention and treatment of atheroscerotic cardiovascuar disease: a statement from the Counci on Cinica Cardioogy (Subcommittee on Exercise, Rehabiitation, and Prevention) and the Counci on Nutrition, Physica Activity, and Metaboism (Subcommittee on Physica Activity). Circuation 2003;107: Hootman JM, Macera CA, Ainsworth BE, Martin M, Addy CL, Bair SN. Association among physica activity eve, cardiorespiratory fitness, and risk of muscuoskeeta injury. Am J Epidemio 2001;154: Ory M, Resnick B, Jordan PJ, Coday M, Riebe D, Ewing Garber C, et a. Screening, safety, and adverse events in physica activity interventions: coaborative experiences from the behavior change consortium. Ann Behav Med 2005;29: Harris T, Kerry SM, Victor CR, Ekeund U, Woodcock A, Iiffe S, et a. A primary care nurse-deivered waking intervention in oder aduts: PACE (pedometer acceerometer consutation evauation)-lift custer randomised controed tria. PLOS Med 2015;12:e /journa.pmed Hisdon M, Foster C, Thorogood M. Interventions for promoting physica activity. Cochrane Database Syst Rev 2005;1:CD Richards J, Thorogood M, Hisdon M, Foster C. Face-to-face versus remote and web 2.0 interventions for promoting physica activity. Cochrane Database Syst Rev 2013;9:CD Foster C, Richards J, Thorogood M, Hisdon M. Remote and web 2.0 interventions for promoting physica activity [pubished onine ahead of print September ]. Cochrane Database Syst Rev King AC, Haske WL, Tayor CB, Kraemer HC, DeBusk RF. Group- vs home-based exercise training in heathy oder men and women. A community-based cinica tria. JAMA 1991;266: Hobbs N, Godfrey A, Lara J, Errington L, Meyer TD, Rochester L, et a. Are behaviora interventions effective in increasing physica activity at 12 to 36 months in aduts aged 55 to 70 years? A systematic review and meta-anaysis. BMC Med 2013;11: King AC, Rejeski WJ, Buchner DM. Physica activity interventions targeting oder aduts. A critica review and recommendations. Am J Prev Med 1998;15: S (98) Richards D. Behaviour change guidance. Evid Based Dent 2007;8: /sj.ebd NICE. Behaviour Change: Individua Approaches. London: NICE; Ogivie D, Foster CE, Rothnie H, Cavi N, Hamiton V, Fitzsimons CF, et a. Interventions to promote waking: systematic review. BMJ 2007;334: Bravata DM, Smith-Spanger C, Sundaram V, Gienger AL, Lin N, Lewis R, et a. Using pedometers to increase physica activity and improve heath: a systematic review. JAMA 2007;298: NIHR Journas Library

147 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Kang M, Marsha SJ, Barreira TV, Lee JO. Effect of pedometer-based physica activity interventions: a meta-anaysis. Res Q Exerc Sport 2009;80: McMurdo ME, Sugden J, Argo I, Boye P, Johnston DW, Sniehotta FF, Donnan PT. Do pedometers increase physica activity in sedentary oder women? A randomized controed tria. JAm Geriatr Soc 2010;58: Houe J, Doyon O, Vadeboncoeur N, Turbide G, Diaz A, Poirier P. Innovative program to increase physica activity foowing an acute coronary syndrome: randomized controed tria. Patient Educ Couns 2011;85:e Yates T, Davies M, Gorey T, Bu F, Khunti K. Effectiveness of a pragmatic education program designed to promote waking activity in individuas with impaired gucose toerance: a randomized controed tria. Diabetes Care 2009;32: NICE. Four Commony Used Methods to Increase Physica Activity: Brief Interventions in Primary Care, Exercise Referra Schemes, Pedometers and Community Based Exercise Programmes for Waking and Cycing. London: NICE; NICE. Waking and Cycing. Loca Measures to Promote Waking and Cycing as Forms of Trave or Recreation. London: NICE; Schneider PL, Bassett DR, Thompson DL, Pronk NP, Bieak KM. Effects of a 10,000 steps per day goa in overweight aduts. Am J Heath Promot 2006;21: Lindberg R. Active iving: on the road with the 10,000 Steps program. J Am Diet Assoc 2000;100: Croteau KA, Richeson NE, Farmer BC, Jones DB. Effect of a pedometer-based intervention on daiy step counts of community-dweing oder aduts. Res Q Exerc Sport 2007;78: Prochaska JJ, Ha SM, Humfeet G, Munoz RF, Reus V, Gorecki J, Hu D. Physica activity as a strategy for maintaining tobacco abstinence: a randomized tria. Prev Med 2008;47: Tudor-Locke C, Craig CL, Brown WJ, Cemes SA, De Cocker K, Gies-Corti B, et a. How many steps/day are enough? For aduts. Int J Behav Nutr Phys Act 2011;8: Marsha SJ, Levy SS, Tudor-Locke CE, Kokhorst FW, Wooten KM, Ji M, et a. Transating physica activity recommendations into a pedometer-based step goa: 3000 steps in 30 minutes. Am J Prev Med 2009;36: McKay J, Wright A, Lowry R, Steee K, Ryde G, Mutrie N. Waking on prescription: the utiity of a pedometer pack for increasing physica activity in primary care. Patient Educ Couns 2009;76: Baker G, Gray SR, Wright A, Fitzsimons C, Nimmo M, Lowry R, et a. The effect of a pedometerbased community waking intervention Waking for Webeing in the West on physica activity eves and heath outcomes: a 12-week randomized controed tria. Int J Behav Nutr Phys Act 2008;5: Tudor-Locke C, Craig CL, Aoyagi Y, Be RC, Croteau KA, De Bourdeaudhuij I, et a. How many steps/day are enough? For oder aduts and specia popuations. Int J Behav Nutr Phys Act 2011;8:80. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 115

148 REFERENCES 47. Macmian F, Fitzsimons C, Back K, Granat MH, Grant MP, Greay M, et a. West End Wakers 65+: a randomised controed tria of a primary care-based waking intervention for oder aduts: study rationae and design. BMC Pubic Heath 2011;11: Mutrie N, Dooin O, Fitzsimons CF, Grant PM, Granat M, Greay M, et a. Increasing oder aduts waking through primary care: resuts of a piot randomized controed tria. Fam Pract 2012;29: Ekeund U, Tingström P, Kamwendo K, Krantz M, Nyander E, Sjöström M, Bergdah B. The vaidity of the Computer Science and Appications activity monitor for use in coronary artery disease patients during eve waking. Cin Physio Funct Imaging 2002;22: /j X x 50. Harris TJ, Owen CG, Victor CR, Adams R, Ekeund U, Cook DG. A comparison of questionnaire, acceerometer, and pedometer: measures in oder peope. Med Sci Sports Exerc 2009;41: Davis MG, Fox KR. Physica activity patterns assessed by acceerometry in oder peope. Eur J App Physio 2007;100: Jefferis BJ, Sartini C, Lee IM, Choi M, Amuzu A, Gutierrez C, et a. Adherence to physica activity guideines in oder aduts, using objectivey measured physica activity in a popuation-based study. BMC Pubic Heath 2014;14: Harris TJ, Owen CG, Victor CR, Adams R, Cook DG. What factors are associated with physica activity in oder peope, assessed objectivey by acceerometry? Br J Sports Med 2009;43: Tudor-Locke C, Ainsworth BE, Thompson RW, Matthews CE. Comparison of pedometer and acceerometer measures of free-iving physica activity. Med Sci Sports Exerc 2002;34: NICE Pubic Heath Coaborating Centre. A Rapid Review of the Effectiveness of Pedometer Interventions to Promote Physica Activity in Aduts. London: NICE; Taib H, Sabirin J. Pedometer and Acceerometer in Assessing Physica Activity Among Chidren and Adoescents. Kuaa Lumpur, Maaysia: Heath Technoogy Assessment Unit; NICE. Modeing the Cost-effectiveness of Physica Activity Interventions. London: NICE; Shaw R, Fenwick E, Baker G, McAdam C, Fitzsimons C, Mutrie N. Pedometers cost buttons : the feasibiity of impementing a pedometer based waking programme within the community. BMC Pubic Heath 2011;11: Heath Checks Programme NHS. Putting Prevention First: NHS Heath Checks: Vascuar Risk Assessment and Management Best Practice Guideines. London: NHS; Lawton BA, Rose SB, Eey CR, Dowe AC, Fenton A, Moyes SA. Exercise on prescription for women aged recruited through primary care: two year randomised controed tria. BMJ 2008;337:a Brawey LR, Rejeski WJ, King AC. Promoting physica activity for oder aduts: the chaenges for changing behavior. Am J Prev Med 2003;25(Supp. 3): S (03)00182-X 62. Tai SS, Goud M, Iiffe S. Promoting heathy exercise among oder peope in genera practice: issues in designing and evauating therapeutic interventions. Br J Gen Pract 1997;47: NIHR Journas Library

149 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Goodman C, Davies S, Tai SS, Dinan S, Iiffe S. Promoting oder peopes participation in activity, whose responsibiity? A case study of the response of heath, oca government and vountary organizations. J Interprof Care 2007;21: Pavey TG, Tayor AH, Fox KR, Hisdon M, Anokye N, Campbe JL, et a. Effect of exercise referra schemes in primary care on physica activity and improving heath outcomes: systematic review and meta-anaysis. BMJ 2011;343:d Khan KM, Weier R, Bair SN. Prescribing exercise in primary care. BMJ 2011;343:d David L. Using CBT in Genera Practice. The 10 Minute Consutation. Boxham, Oxfordshire: Scion; British Psychoogica Society. Improving Heath: Changing Behaviour: NHS Heath Trainer Handbook. London: Department of Heath and Socia Care; Michie S, Ashford S, Sniehotta FF, Dombrowski SU, Bishop A, French DP. A refined taxonomy of behaviour change techniques to hep peope change their physica activity and heathy eating behaviours: the CALO-RE taxonomy. Psycho Heath 2011;26: Harris T, Kerry S, Victor C, Ekeund U, Woodcock A, Iiffe S, et a. Randomised controed tria of a compex intervention by primary care nurses to increase waking in patients aged years: protoco of the PACE-Lift (Pedometer Acceerometer Consutation Evauation - Lift) tria. BMC Pubic Heath 2013;13: Harris T, Kerry SM, Victor CR, Shah SM, Iiffe S, Ussher M, et a. PACE-UP (Pedometer and consutation evauation UP) a pedometer-based waking intervention with and without practice nurse support in primary care patients aged years: study protoco for a randomised controed tria. Trias 2013;14: Department for Communities and Loca Government. Engish Indices of Deprivation 2010: Guidance Document. London: Department for Communities and Loca Government; Booth M. Assessment of physica activity: an internationa perspective. Res Q Exerc Sport 2000;71(Supp. 2): Physica Activity Poicy Unit, Department of Heath and Socia Care. The Genera Practice Physica Activity Questionnaire (GPPAQ). London: Department of Heath and Socia Care; Jette AM, Rooks D, Lachman M, Lin TH, Levenson C, Heisein D, et a. Home-based resistance training: predictors of participation and adherence. Gerontoogist 1998;38: Zigmond AS, Snaith RP. The Hospita Anxiety and Depression Scae. Acta Psychiatr Scand 1983;67: Brooks R. EuroQo: the current state of pay. Heath Poicy 1996;37: / (96) Ware JE, Sherbourne CD. The MOS 36-item short-form heath survey (SF-36). I. Conceptua framework and item seection. Med Care 1992;30: Ahmad S, Harris T, Limb E, Kerry S, Victor C, Ekeund U, et a. Evauation of reiabiity and vaidity of the Genera Practice Physica Activity Questionnaire (GPPAQ) in year od primary care patients. BMC Fam Pract 2015;16: Office for Nationa Statistics. The Engish and Wesh 2011 Census. Newport: Office for Nationa Statistics; Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 117

150 REFERENCES 80. McGee MA, Johnson AL, Kay DW. The description of activities of daiy iving in five centres in Engand and Waes. Medica Research Counci Cognitive Function and Ageing Study. Age Ageing 1998;27: Tunsta J. Od and Aone. London: Routedge and Kegan; Nandy S, Parsons S, Cryer C, Underwood M, Rashbrook E, Carter Y, et a. Deveopment and preiminary examination of the predictive vaidity of the Fas Risk Assessment Too (FRAT) for use in primary care. J Pubic Heath 2004;26: Caspersen CJ, Boemberg BP, Saris WH, Merritt RK, Kromhout D. The prevaence of seected physica activities and their reation with coronary heart disease risk factors in edery men: the Zutphen Study, Am J Epidemio 1991;133: Freedson PS, Meanson E, Sirard J. Caibration of the Computer Science and Appications, Inc. acceerometer. Med Sci Sports Exerc 1998;30: Matthews CE, Chen KY, Freedson PS, Buchowski MS, Beech BM, Pate RR, Troiano RP. Amount of time spent in sedentary behaviors in the United States, Am J Epidemio 2008;167: Litte P, Dorward M, Graton S, Hammerton L, Piinger J, White P, et a. A randomised controed tria of three pragmatic approaches to initiate increased physica activity in sedentary patients with risk factors for cardiovascuar disease. Br J Gen Pract 2004;54: Stevens W, Hisdon M, Thorogood M, McArde D. Cost-effectiveness of a primary care based physica activity intervention in year od men and women: a randomised controed tria. Br J Sports Med 1998;32: Munro JF, Nicho JP, Brazier JE, Davey R, Cochrane T. Cost effectiveness of a community based exercise programme in over 65 year ods: custer randomised tria. J Epidemio Community Heath 2004;58: Sugden JA, Sniehotta FF, Donnan PT, Boye P, Johnston DW, McMurdo ME. The feasibiity of using pedometers and brief advice to increase activity in sedentary oder women a piot study. BMC Heath Serv Res 2008;8: Trost SG, McIver KL, Pate RR. Conducting acceerometer-based activity assessments in fied-based research. Med Sci Sports Exerc 2005;37(Supp. 11): mss Mier GD, Jakicic JM, Rejeski WJ, Whit-Gover MC, Lang W, Wakup MP, Hodges ML. Effect of varying acceerometry criteria on physica activity: the ook ahead study. Obesity 2013;21: Senn S. Change from baseine and anaysis of covariance revisited. Stat Med 2006;25: Harris T, Kerry SM, Limb ES, Victor CR, Iiffe S, Ussher M, et a. Effect of a primary care waking intervention with and without nurse support on physica activity eves in 45- to 75-year-ods: the Pedometer And Consutation Evauation (PACE-UP) custer randomised cinica tria. PLOS Med 2017;14:e Department of Heath and Socia Care. Heathy Lives, Heathy Peope: Our Strategy for Pubic Heath in Engand. London: Department of Heath and Socia Care; Department of Heath and Socia Care. Improving Outcomes: A Strategy for Cancer. London: Department of Heath and Socia Care; NIHR Journas Library

151 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Fitzsimons CF, Baker G, Wright A, Nimmo MA, Ward Thompson C, Lowry R, et a. The Waking for Webeing in the West randomised controed tria of a pedometer-based waking programme in combination with physica activity consutation with 12 month foow-up: rationae and study design. BMC Pubic Heath 2008;8: Leung W, Ashton T, Kot GS, Schofied GM, Garrett N, Kerse N, Pate A. Cost-effectiveness of pedometer-based versus time-based Green Prescriptions: the Heathy Steps Study. Aust J Prim Heath 2012;18: Over EA, Wende-Vos GW, van den Berg M, Reenen HH, Tariq L, Hoogenveen RT, van Baa PH. Cost-effectiveness of counseing and pedometer use to increase physica activity in the Netherands: a modeing study. Cost Eff Resour Aoc 2012;10: Cobiac LJ, Vos T, Barendregt JJ. Cost-effectiveness of interventions to promote physica activity: a modeing study. PLOS Med 2009;6:e Vijay GC, Wison ECF, Suhrcke M, Hardeman W, Sutton S. Are brief interventions to increase physica activity cost-effective? Br J Sports Med 2016;50: bjsports Drummond M SM, Kaxton K, Stoddart G, Torrance G, editors. Methods for the Economic Evauation of Heath Care Programmes. 4th edn. Oxford: Oxford University Press; Department of Heath and Socia Care. NHS Reference Costs London: Department of Heath and Socia Care; Curtis LA. Unit Costs of Heath and Socia Care Canterbury: PSSRU, University of Kent; Anokye NK, Pokhre S, Fox-Rushby J. Economic anaysis of participation in physica activity in Engand: impications for heath poicy. Int J Behav Nutr Phys Act 2014;11: /s Feng Y, Devin N, Herdman M. Assessing the heath of the genera popuation in Engand: how do the three- and five-eve versions of EQ-5D compare? Heath Qua Life Outcomes 2015;13: Rabin R, Oemar M, Oppe M, Janssen B, Herdman M. EQ-5D-5L User Guide: Basic Information on How to Use the EQ-5D-5L Instrument. Rotterdam: EuroQo Group; Doan P. Modeing vauations for EuroQo heath states. Med Care 1997;35: Lydick E, Epstein RS, Himmeberger D, White CJ. Area under the curve: a metric for patient subjective responses in episodic diseases. Qua Life Res 1995;4: /BF Briggs A, Cark T, Wostenhome J, Carke P. Missing... presumed at random: cost-anaysis of incompete data. Heath Econ 2003;12: Schafer JL. Mutipe imputation: a primer. Stat Methods Med Res 1999;8: / Bennett DA. How can I dea with missing data in my study? Aust N Z J Pubic Heath 2001;25: Simons CL, Rivero-Arias O, Yu LM, Simon J. Mutipe imputation to dea with missing EQ-5D-3L data: shoud we impute individua domains or the actua index? Qua Life Res 2015;24: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 119

152 REFERENCES 113. Rubin D. Mutipe Imputation for Nonresponse in Surveys United States of America. Hoboken, NJ: John Wiey & Sons; Aison PD. Mutipe imputation for missing data: a cautionary tae. Socio Method Res 2000;28: Carpenter JR, Kenward MG. MAR Methods for Quantitative Data. In Carpenter JR, Kenward MG, editors. Missing Data In Randomised Controed Trias: A Practica Guide. Birmingham: Heath Technoogy Assessment Methodoogy Programme; pp Sterne JA, White IR, Carin JB, Spratt M, Royston P, Kenward MG, et a. Mutipe imputation for missing data in epidemioogica and cinica research: potentia and pitfas. BMJ 2009;338:b Gick HA, Doshi JA, Sonnad SS, Posky D. Economic Evauation in Cinica Trias. 2nd edn. Oxford: Oxford University Press; Vu T, Harris A, Duncan G, Sussman G. Cost-effectiveness of mutidiscipinary wound care in nursing homes: a pseudo-randomized pragmatic custer tria. Fam Pract 2007;24: Basu A, Manca A. Regression estimators for generic heath-reated quaity of ife and quaity-adjusted ife years. Med Decis Making 2012;32: Manning WG, Muahy J. Estimating og modes: to transform or not to transform? J Heath Econ 2001;20: Smith L, Gardner B, Fisher A, Hamer M. Patterns and correates of physica activity behaviour over 10 years in oder aduts: prospective anayses from the Engish Longitudina Study of Ageing. BMJ Open 2015;5:e Henderson C, Knapp M, Fernández JL, Beecham J, Hirani SP, Cartwright M, et a. Cost effectiveness of teeheath for patients with ong term conditions (Whoe Systems Demonstrator teeheath questionnaire study): nested economic evauation in a pragmatic, custer randomised controed tria. BMJ 2013;346:f Graubard BI, Korn EL. Anayzing heath surveys for cancer-reated objectives. J Nat Cancer Inst 1999;91: Reiner M, Niermann C, Jekauc D, Wo A. Long-term heath benefits of physica activity a systematic review of ongitudina studies. BMC Pubic Heath 2013;13: / Sattemair J, Pertman J, Ding EL, Koh HW, Haske W, Lee IM. Dose response between physica activity and risk of coronary heart disease: a meta-anaysis. Circuation 2011;124: Jeon CY, Lokken RP, Hu FB, van Dam RM. Physica activity of moderate intensity and risk of type 2 diabetes: a systematic review. Diabetes Care 2007;30: Warburton DE, Nico CW, Bredin SS. Heath benefits of physica activity: the evidence. CMAJ 2006;174: Lee CD, Fosom AR, Bair SN. Physica activity and stroke risk: a meta-anaysis. Stroke 2003;34: Oanrewaju O, Key S, Cowan A, Brayne C, Lafortune L. Physica activity in community dweing oder peope: a systematic review of reviews of interventions and context. PLOS ONE 2016;11:e NIHR Journas Library

153 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Anokye NK, Lord J, Fox-Rushby J. Is brief advice in primary care a cost-effective way to promote physica activity? Br J Sports Med 2014;48: Campbe F, Homes M, Everson-Hock E, Davis S, Buckey Woods H, Anokye N, et a. A systematic review and economic evauation of exercise referra schemes in primary care: a short report. Heath Techno Assess 2015;19(60) Pavey TG, Anokye N, Tayor AH, Trueman P, Moxham T, Fox KR, et a. The cinica effectiveness and cost-effectiveness of exercise referra schemes: a systematic review and economic evauation. Heath Techno Assess 2011;15(44) Anokye NK, Trueman P, Green C, Pavey TG, Hisdon M, Tayor RS. The cost-effectiveness of exercise referra schemes. BMC Pubic Heath 2011;11: Anokye N, Anagnostou A, Lord J, Vai Y, Tayor S, Foster C, et a. An Individua-eve Simuation Mode for Assessing Physica Activity Interventions. European Pubic Heath Conference, Vienna, NICE. Physica Activity: Brief Advice for Aduts in Primary Care. Pubic heath guideine [PH44]. London: NICE; Matrix. Estimating Return on Investment for Interventions and Strategies to Increase Physica Activity-Technica Report. London: NICE; Office for Nationa Statistics. Life Expectancy URL: (accessed 27 November 2017) Physica Activity Guideines Advisory Committee. Physica Activity Guideines Advisory Committee Report, Washington, DC: Department of Heath and Human Services; Laine J, Kuvaja-Köner V, Pietiä E, Koivuneva M, Vatonen H, Kankaanpää E. Cost-effectiveness of popuation-eve physica activity interventions: a systematic review. Am J Heath Promot 2014;29: Garrett S, Eey CR, Rose SB, O Dea D, Lawton BA, Dowe AC. Are physica activity interventions in primary care and the community cost-effective? A systematic review of the evidence. Br J Gen Pract 2011;61:e De Smedt D, De Cocker K, Annemans L, De Bourdeaudhuij I, Cardon G. A cost-effectiveness study of the community-based intervention Steps Ghent. Pubic Heath Nutr 2012;15: Brennan A, Bake L, Hi-McManus D, Payne N, Buckey Woods H, Bank L. Waking and Cycing: Loca Measures to Promote Waking and Cycing As Forms of Trave or Recreation: Heath Economic and Modeing Report. London: NICE; NHS Digita. Heath Survey for Engand 2004: The Heath of Ethnic Minorities, Headine Resuts. London: NHS Digita; Lamb SE, Bartett HP, Ashey A, Bird W. Can ay-ed waking programmes increase physica activity in midde aged aduts? A randomised controed tria. J Epidemio Community Heath 2002;56: Copeand RJ, Horspoo K, Humphreys L, Scott E, Booster tria team. Recruiting to a arge-scae physica activity randomised controed tria - experiences with the gift of hindsight. Trias 2016;17: Caperchione CM, Duncan M, Kot GS, Vandeanotte C, Rosenkranz RR, Maeder A, et a. Examining an Austraian physica activity and nutrition intervention using RE-AIM. Heath Promot Int 2016;31: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 121

154 REFERENCES 147. Kerry SM, Morgan KE, Limb E, Cook DG, Furness C, Carey I, et a. Interpreting popuation reach of a arge, successfu physica activity tria deivered through primary care. BMC Pubic Heath 2018;18: Normanse R, Homes R, Victor C, Cook DG, Kerry S, Iiffe S, et a. Exporing non-participation in primary care physica activity interventions: PACE-UP tria interview findings. Trias 2016;17: Victor CR, Rogers A, Woodcock A, Beighton C, Cook DG, Kerry SM, et a. What factors support oder peope to increase their physica activity eves? An exporatory anaysis of the experiences of PACE-Lift tria participants. Arch Geronto Geriatr 2016;67: j.archger Ritchie J, Spencer L. Quaitative Data Anaysis for Appied Poicy Research. In Bryman AB, Burgess RG, editors. Anaysing Quaitative Data. London: Routedge; pp Foster CE, Brennan G, Matthews A, McAdam C, Fitzsimons C, Mutrie N. Recruiting participants to waking intervention studies: a systematic review. Int J Behav Nutr Phys Act 2011;8: Loudon K, Treweek S, Suivan F, Donnan P, Thorpe KE, Zwarenstein M. The PRECIS-2 too: designing trias that are fit for purpose. BMJ 2015;350:h Warner ET, Gasgow RE, Emmons KM, Bennett GG, Askew S, Rosner B, Coditz GA. Recruitment and retention of participants in a pragmatic randomized intervention tria at three community heath cinics: resuts and essons earnt. BMC Pubic Heath 2013;13: Dubbert PM, Cooper KM, Kirchner KA, Meydrech EF, Bibrew D. Effects of nurse counseing on waking for exercise in edery primary care patients. J Geronto A Bio Sci Med Sci 2002;57:M Dubbert PM, Morey MC, Kirchner KA, Meydrech EF, Grothe K. Counseing for home-based waking and strength exercise in oder primary care patients. Arch Intern Med 2008;168: Cappuccio FP, Cook DG, Atkinson RW, Wicks PD. The Wandsworth Heart and Stroke Study. A popuation-based survey of cardiovascuar risk factors in different ethnic groups. Methods and baseine findings. Nutr Metab Cardiovasc Dis 1998;8: Rogers A, Harris T, Victor C, Woodcock A, Limb E, Kerry S, et a. Which oder peope decine participation in a primary care tria of physica activity and why: insights from a mixed methods approach. BMC Geriatr 2014;14: Attwood S, Morton K, Sutton S. Exporing equity in uptake of the NHS Heath Check and a nested physica activity intervention tria. J Pubic Heath (Oxf) 2016;38: /pubmed/fdv Dougas A, Bhopa RS, Bhopa R, Forbes JF, Gi JM, Lawton J, et a. Recruiting South Asians to a ifestye intervention tria: experiences and essons from PODOSA (Prevention of Diabetes & Obesity in South Asians). Trias 2011;12: Wibur J, Buchhoz SW, Ingram DM, Braun LT, Johnson TJ, Fogg L, et a. Effectiveness, efficiency, duration, and costs of recruiting for an African American women s ifestye physica activity program. Res Nurs Heath 2013;36: Crombie IK, Irvine L, Wiiams B, McGinnis AR, Sane PW, Ader EM, et a. Why oder peope do not participate in eisure time physica activity: a survey of activity eves, beiefs and deterrents. Age Ageing 2004;33: NIHR Journas Library

155 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Normanse R, Smith J, Victor C, Cook DG, Kerry S, Iiffe S, et a. Numbers are not the whoe story: a quaitative exporation of barriers and faciitators to increased physica activity in a primary care based waking intervention. BMC Pubic Heath 2014;14: Chinn DJ, White M, Howe D, Harand JO, Drinkwater CK. Factors associated with non-participation in a physica activity promotion tria. Pubic Heath 2006;120: j.puhe Costeo E, Kafchinski M, Vraze J, Suivan P. Motivators, barriers, and beiefs regarding physica activity in an oder adut popuation. J Geriatr Phys Ther 2011;34: JPT.0b013e31820e0e Cohen-Mansfied J, Marx MS, Guranik JM. Motivators and barriers to exercise in an oder community dweing popuation. J Ageing Phys Act 2003;11: japa Justine M, Azizan A, Hassan V, Saeh Z, Manaf H. Barriers to participation in physica activity and exercise among midde-aged and edery individuas. Singapore Med J 2013;54: Booth ML, Bauman A, Owen N, Gore CJ. Physica activity preferences, preferred sources of assistance, and perceived barriers to increased activity among physicay inactive Austraians. Prev Med 1997;26: Bock C, Jarczok MN, Litaker D. Community-based efforts to promote physica activity: a systematic review of interventions considering mode of deivery, study quaity and popuation subgroups. J Sci Med Sport 2014;17: Moore GF, Audrey S, Barker M, Bond L, Bone C, Hardeman W, et a. Process evauation of compex interventions: Medica Research Counci guidance. BMJ 2015;350:h /bmj.h Furness C, Howard E, Limb E, Cook DG, Kerry S, Wahich C, et a. Reating process evauation measures to compex intervention outcomes: findings from the PACE-UP primary care pedometerbased waking tria. Trias 2018;19: Horvath AO, Greenberg LS. Deveopment and vaidation of the working aiance inventory. J Couns Psycho 1989;36: Andrusyna TP, Tang TZ, DeRubeis RJ, Luborsky L. The factor structure of the working aiance inventory in cognitive-behaviora therapy. J Psychother Pract Res 2001;10: Low DC, Mier SD, Squire B. The Outcome Rating Scae (ORS) and Session Rating Scaes (SRS): Feedback Informed Treatment in Chid and Adoescent Menta Heath Services (CAMHS) URL: paper%20for%20cyp-iapt.pdf (accessed 18 June 2018) Mannix KA, Backburn IM, Garand A, Gracie J, Moorey S, Reid B, et a. Effectiveness of brief training in cognitive behaviour therapy techniques for paiative care practitioners. Paiat Med 2006;20: Hope R. The Ten Essentia Shared Capabiities A Framework for the Whoe of the Menta Heath Workforce. London: Department of Heath and Socia Care; e Ramsay CR, Thomas RE, Croa BL, Grimshaw JM, Ecces MP. Using the theory of panned behaviour as a process evauation too in randomised trias of knowedge transation strategies: a case study from UK primary care. Impement Sci 2010;5: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 123

156 REFERENCES 177. Carke DJ, Godfrey M, Hawkins R, Sader E, Harding G, Forster A, et a. Impementing a training intervention to support caregivers after stroke: a process evauation examining the initiation and embedding of programme change. Impement Sci 2013;8: Tomasone JR, Martin Ginis KA, Estabrooks PA, Domenicucci L. Changing minds, changing ives from the top down: an investigation of the dissemination and adoption of a Canada-wide educationa intervention to enhance heath care professionas intentions to prescribe physica activity. Int J Behav Med 2015;22: Tomasone JR, Arbour-Nicitopouos KP, Pia E, Lamontagne ME, Cummings I, Latimer-Cheung AE, Routhier F. Exporing end user adoption and maintenance of a teephone-based physica activity counseing service for individuas with physica disabiities using the theoretica domains framework. Disabi Rehabi 2017;39: Hasson H, Bomberg S, Dunér A. Fideity and moderating factors in compex interventions: a case study of a continuum of care program for frai edery peope in heath and socia care. Impement Sci 2012;7: Grimshaw JM, Zwarenstein M, Tetroe JM, Godin G, Graham ID, Lemyre L, et a. Looking inside the back box: a theory-based process evauation aongside a randomised controed tria of printed educationa materias (the Ontario printed educationa message, OPEM) to improve referra and prescribing practices in primary care in Ontario, Canada. Impement Sci 2007;2: Fitzgerad TM, Wiiams PA, Dodge JA, Quinn M, Heminger CL, Moutrie R, et a. Program impementation approaches to buid and sustain heath care coordination for type 2 diabetes. Heath Promot Pract 2017;18: Campbe-Scherer DL, Assein J, Osunana AM, Fieding S, Anderson R, Rueda-Causen CF, et a. Impementation and evauation of the 5As framework of obesity management in primary care: design of the 5As Team (5AsT) randomized contro tria. Impement Sci 2014;9: / Berendsen BA, Kremers SP, Saveberg HH, Schaper NC, Hendriks MR. The impementation and sustainabiity of a combined ifestye intervention in primary care: mixed method process evauation. BMC Fam Pract 2015;16: Beighton C, Victor C, Normanse R, Cook D, Kerry S, Iiffe S, et a. It s not just about waking... it s the practice nurse that makes it work : a quaitative exporation of the views of practice nurses deivering compex physica activity interventions in primary care. BMC Pubic Heath 2015;15: Picorei AM, Pereira LS, Pereira DS, Feício D, Sherrington C. Adherence to exercise programs for oder peope is infuenced by program characteristics and persona factors: a systematic review. J Physiother 2014;60: Wahich C, Beighton C, Victor C, Normanse R, Cook D, Kerry S, et a. You started something... then I continued by mysef : a quaitative study of physica activity maintenance. Prim Heath Care Res Dev 2017;18: Harris T, Kerry SM, Limb ES, Furness C, Wahich C, Victor CR, et a. Physica activity eves in aduts and oder aduts 3-4 years after pedometer-based waking interventions: ong-term foow-up of participants from two randomised controed trias in UK primary care. PLOS Med 2018;15:e NIHR Journas Library

157 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Iiffe S, Kendrick D, Morris R, Masud T, Gage H, Sketon D, et a. Muticentre custer randomised tria comparing a community group exercise programme and home-based exercise with usua care for peope aged 65 years and over in primary care. Heath Techno Assess 2014;18(49) Marcus BH, Dubbert PM, Forsyth LH, McKenzie TL, Stone EJ, Dunn AL, Bair SN. Physica activity behavior change: issues in adoption and maintenance. Heath Psycho 2000;19: White IR, Carpenter J, Horton N. A Mean Score Method for Sensitivity Anaysis to Departures from the Missing at Random Assumption in Randomised Trias. Statistica Sinica; URL: (accessed 25 May 2018) Tuy MA, Cuppes ME, Chan WS, McGade K, Young IS. Brisk waking, fitness, and cardiovascuar risk: a randomized controed tria in primary care. Prev Med 2005;41: /j.ypmed NHS Engand Pubic Heath Engand Diabetes UK. Heathier You: NHS Diabetes Prevention Programme (NHS DPP) URL: diabetes-prevention (accessed 27 November 2017) Chau JY, Grunseit AC, Chey T, Stamatakis E, Brown WJ, Matthews CE, et a. Daiy sitting time and a-cause mortaity: a meta-anaysis. PLOS ONE 2013;8:e journa.pone Biswas A, Oh PI, Faukner GE, Bajaj RR, Siver MA, Mitche MS, Ater DA. Sedentary time and its association with risk for disease incidence, mortaity, and hospitaization in aduts: a systematic review and meta-anaysis. Ann Intern Med 2015;162: Fitzsimons CF, Baker G, Gray SR, Nimmo MA, Mutrie N, Scottish Physica Activity Research Coaboration (SPARCo). Does physica activity counseing enhance the effects of a pedometerbased intervention over the ong-term: 12-month findings from the Waking for Webeing in the west study. BMC Pubic Heath 2012;12: Zheng H, Orsini N, Amin J, Wok A, Nguyen VT, Ehrich F. Quantifying the dose-response of waking in reducing coronary heart disease risk: meta-anaysis. Eur J Epidemio 2009;24: Dwyer T, Pezic A, Sun C, Cochrane J, Venn A, Srikanth V, et a. Objectivey measured daiy steps and subsequent ong term a-cause mortaity: the tasped prospective cohort study. PLOS ONE 2015;10:e Marcus BH, Lewis BA, Hogan J, King TK, Abrecht AE, Bock B, et a. The efficacy of moderateintensity exercise as an aid for smoking cessation in women: a randomized controed tria. Nicotine Tob Res 2005;7: Ashford S, Edmunds J, French DP. What is the best way to change sef-efficacy to promote ifestye and recreationa physica activity? A systematic review with meta-anaysis. Br J Heath Psycho 2010;15: Cox KL, Ficker L, Ameida OP, Xiao J, Greenop KR, Hendriks J, et a. The FABS tria: a randomised contro tria of the effects of a 6-month physica activity intervention on adherence and ong-term physica activity and sef-efficacy in oder aduts with memory compaints. Prev Med 2013;57: Kuer LH, Kinze LS, Pettee KK, Kriska AM, Simkin-Siverman LR, Conroy MB, et a. Lifestye intervention and coronary heart disease risk factor changes over 18 months in postmenopausa women: the Women On the Move through Activity and Nutrition (WOMAN study) cinica tria. J Womens Heath 2006;15: Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 125

158 REFERENCES 203. Opdenacker J, Boen F, Coorevits N, Deecuse C. Effectiveness of a ifestye intervention and a structured exercise intervention in oder aduts. Prev Med 2008;46: /j.ypmed Tuoch H, Sweet SN, Fortier M, Capstick G, Kenny GP, Siga RJ. Exercise faciitators and barriers from adoption to maintenance in the diabetes aerobic and resistance exercise tria. Can J Diabetes 2013;37: Kaewthummanuku T, Brown KC. Determinants of empoyee participation in physica activity: critica review of the iterature. AAOHN J 2006;54: Sharma M, Sargent L, Stacy R. Predictors of eisure-time physica activity among African American women. Am J Heath Behav 2005;29: Office for Nationa Statistics (ONS). Annua survey of hours and earnings. ONS Digita; URL: buetins/annuasurveyofhoursandearnings/2015provisionaresuts (accessed November 2015) Sharpes L, Gover M, Bennett M, Jordan J, Cutterbuck-James A, Davies M, et a. TOMADO: a crossover randomised controed tria of ora mandibuar advancement devices for obstructive seep apnoea-hypopnoea. Heath Techno Assess 2014;18(67) Giett M, Daosso HM, Dixon S, Brennan A, Carey ME, Campbe MJ, et a. Deivering the diabetes education and sef-management for ongoing and newy diagnosed (DESMOND) programme for peope with newy diagnosed type 2 diabetes: cost effectiveness anaysis. BMJ 2010;341:c Richardson G, Kennedy A, Reeves D, Bower P, Lee V, Middeton E, et a. A Cost effectiveness of the Expert Participants Programme (EPP) for participants with chronic conditions. J Epidemio and Community Heath 2008;62: Hu G, Qiao Q, Siventoinen K, Eriksson JG, Jousiahti P, Lindström J, et a. Occupationa, commuting, and eisure-time physica activity in reation to risk for Type 2 diabetes in midde-aged Finnish men and women. Diabetoogia 2003;46: Hu G, Sarti C, Jousiahti P, Siventoinen K, Barengo NC, Tuomiehto J. Leisure time, occupationa, and commuting physica activity and the risk of stroke. Stroke 2005;36: /01.STR c 213. Hu G, Jousiahti P, Boroduin K, Barengo NC, Lakka TA, Nissinen A, Tuomiehto J. Occupationa, commuting and eisure-time physica activity in reation to coronary heart disease among midde-aged Finnish men and women. Atheroscerosis 2007;194: j.atheroscerosis Briggs AS, Scupher MJ, Caxton K. Decision Modeing for Heath Economic Evauation. Oxford: Oxford University Press; Hu G, Jousiahti P, Antikainen R, Tuomiehto J. Occupationa, commuting, and eisure-time physica activity in reation to cardiovascuar mortaity among Finnish subjects with hypertension. Am J Hypertens 2007;20: Preis SR, Pencina MJ, Hwang SJ, D Agostino RB, Savage PJ, Levy D, Fox CS. Trends in cardiovascuar disease risk factors in individuas with and without diabetes meitus in the Framingham Heart Study. Circuation 2009;120: Ward SL, Loyd Jones M, Pandor A, Homes M, Ara R, Ryan A. et a. A systematic review and economic evauation of statins for the prevention of coronary. Heath Techno Assess 2007;11(14). 126 NIHR Journas Library

159 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Nationa Cinica Guideine Centre (UK). Hypertension: The Cinica Management of Primary Hypertension in Aduts: Update of Cinica Guideines 18 and 34. NICE Cinica Guideines. London: NICE; Jamrozik K. Estimate of deaths attributabe to passive smoking among UK aduts: database anaysis. BMJ 2005;330: Brønnum-Hansen H, Davidsen M, Thorvadsen P, Danish MONICA Study Group. Long-term surviva and causes of death after stroke. Stroke 2001;32: hs Preis SR, Hwang SJ, Coady S, Pencina MJ, D Agostino RB Sr, Savage PJ, et a. Trends in a-cause and cardiovascuar disease mortaity among women and men with and without diabetes meitus in the Framingham Heart Study, 1950 to Circuation 2009;119: /CIRCULATIONAHA Ward S, Loyd Jones M, Pandor A, Homes M, Ara R, Ryan A, et a. Statins for the prevention of coronary events. Technoogy assessment report commissioned by the HTA programme on behaf of The Nationa Institute for Cinica Exceence. London: NICE; Nationa Cinica Guideine Centre (NCGC). Hypertension: the cinica management of primary hypertension in aduts. Cinica guideine: methods, evidence and recommendations. London: NICE; Gonzaez ELM, Johansson S, Waander MA. Trends in the prevaence and incidence of diabetes in the UK, J Epidemio Community Heath 2009;63: jech Department for Heath and Socia Care. Heath Survey for Engand 2011, Heath, Socia Care and Lifestyes. London: Department for Heath and Socia Care; Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 127

160

161 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Appendix 1 Methods study Heath and ifestye survey Study IDNO Thank you for fiing in this questionnaire. It wi take you about minutes to compete. Pease fee free to write comments by any question. A information wi be kept stricty confidentia. Pease enter your date of birth Pease enter today s date / / / / Thank you Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 129

162 APPENDIX 1 Section A - Some genera questions about your heath Pease put a tick in the box next to the most appropriate answer for each question. How is your heath in genera? Very good Good Fair Poor Very poor Are your day-to-day activities imited because of a heath probem or disabiity which has asted, or is expected to ast, at east 12 months? (Incude probems reated to od age.) Yes, imited a ot Yes, imited a itte No 3 How much physica or bodiy pain have you had in the past 4 weeks? None Very mid or mid Moderate Severe or very severe 4 In the past four weeks, how much did pain interfere with your norma activities? Not at a A itte bit Moderatey Quite a bit or extremey 130 NIHR Journas Library

163 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section B - specific questions about your heath Have you ever been tod by a doctor or nurse that you have any of these conditions? (Pease tick a that appy to you) YES Angina.... A heart attack Other heart probems Stroke... High bood pressure... Chronic bronchitis.... Asthma..... Diabetes.... Arthritis Cancer (apart from skin cancer).... Depression..... Parkinson s Disease How many times have you faen over in the ast year? None Once or twice Three times or more Not sure 14 How many different medications do you take every day? None One Two Three Four or more 15 Have you ever smoked? Yes No (pease go to question 17) 16 Do you currenty smoke? Yes No 17 One unit of acoho is approximatey haf a pint of beer / cider, one gass of wine or sherry, or a singe whisky, gin etc. Approximatey how many units of acoho do you have during the average week?... units. Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 131

164 APPENDIX 1 Section C - Questions about your heath today Under each heading, pease tick the ONE box that best describes your heath TODAY 1 Mobiity I have no probems in waking about I have sight probems in waking about I have moderate probems in waking about I have severe probems in waking about I am unabe to wak about 2 Sef-care I have no probems with sef-care I have sight probems washing or dressing mysef I have moderate probems washing or dressing mysef I have severe probems washing or dressing mysef I am unabe to wash or dress mysef 3 Usua activities (e.g. work, study, housework, famiy or eisure) I have no probems doing my usua activities I have sight probems doing my usua activities I have moderate probems doing my usua activities I have severe probems doing my usua activities I am unabe to do my usua activities 4 Pain / discomfort I have no pain or discomfort I have sight pain or discomfort I have moderate pain or discomfort I have severe pain or discomfort I have extreme pain or discomfort 5 Anxiety / depression I am not anxious or depressed I am sighty anxious or depressed I am moderatey anxious or depressed I am severey anxious or depressed I am extremey anxious or depressed 132 NIHR Journas Library

165 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section C - Your heath today (continued) We woud ike to know how good or bad your heath is TODAY The scae is numbered 0 to means the best heath you can imagine 0 means the worst heath you can imagine Mark an X on the scae to indicate how your heath is TODAY Now, pease write the number you marked on the scae in the box beow YOUR HEALTH TODAY = Section D - Your contact with your GP surgery 1. During the ast 3 months did you tak to a doctor or nurse at your genera practice on your own behaf, either in person or by teephone? Yes No (If no, pease go to section E) If yes, approximatey how many times did this happen in the ast 3 months? Once Twice Three times Four or more times Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 133

166 APPENDIX 1 Section E - Some questions on how you fee For each item beow, pease tick the box opposite the repy that comes cosest to how you have been feeing over the past week. Don t take too ong over the answers: your immediate reaction wi probaby be most accurate. Tick ony one box in each section 1. I fee tense or wound up : Most of the time A ot of the time From time to time Not at a 2. I fee as if I am sowed down: Neary a of the time Very often Sometimes Not at a 3. I sti enjoy things I used to: Definitey as much Not quite as much Ony a itte Hardy at a 4. I get a sort of frightened feeing ike butterfies in the stomach: Not at a Occasionay Quite often Very often 5. I get a sort of frightened feeing as if something bad is about to happen: Very definitey Yes, but not too bady A itte, but it doesn t worry me Not at a 6. I have ost interest in my appearance: Definitey I don t take so much care as I shoud do I might not take quite as much care I take just as much care 7. I can augh and see the funny side of things: As much as I aways coud Not quite so much now Definitey not so much now Not at a 134 NIHR Journas Library

167 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO I fee restess, as if I have to be on the move Very much indeed Quite a ot Not very much Not at a 9. Worrying thoughts go through my mind: A great dea of the time A ot of the time From time to time but not too often Ony occasionay 10. I ook forward with enjoyment to things: As much as I ever did Rather ess than I used to Definitey ess than I used to Hardy at a 11. I fee cheerfu: Not at a Not often Sometimes Most of the time 12. I get sudden feeings of panic Very often indeed Quite often Not very often Not at a 13. I can sit at ease and fee reaxed: Definitey Usuay Not often Not at a 14. I can enjoy a good book, radio or TV programme: Often Sometimes Not often Very sedom 15. I fee oney: A the time Often Sometimes Never Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 135

168 APPENDIX 1 Section F - Some questions about difficuties you may have Here are a few things some peope find difficut to do without hep. Do you or woud you have difficuty with these activities? No Some Unabe Difficuty Difficuty to do aone 1 Washing yoursef a over 2 Cutting your own toenais 3 Getting on a bus 4 Going up and down stairs 5 Doing heavy housework 6 Shopping & carrying heavy bags 7 Preparing and cooking a hot mea 8 Reaching an overhead shef 9 Tying a good knot in a piece of string 10 Do you have any probems with your baance? No Yes 11 Can you see we enough to recognise a friend across a road? Yes, without gasses Yes, with gasses No 136 NIHR Journas Library

169 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section G- Some questions about your attitudes to exercise and heath Pease tick one box to indicate how strongy you agree or disagree with each statement Strongy agree Sighty agree Unsure Sighty disagree Strongy disagree 1. Doing exercise is satisfying and rewarding to me 2. Doing exercise reguary is good for me 3. There is itte I can do to make up for the physica osses that come with age 4. Exercising reguary can be hepfu for my heath 5. Exercising reguary can hep me to get out of doors 6. Exercising reguary can hep me to contro my weight or to ose weight Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 137

170 APPENDIX 1 Section H - Some questions about physica activity 1 How many times did you take a wak outside during the ast week? (incude waking reated to other activities e.g. for shopping, trave to work etc)...times ast week 2 How ong did such a wak usuay ast?..minutes 3 Did you take a wak that asted onger than 1 hour during the ast month? Yes No 3a If yes, how many times did you do that?.times in the ast month 4 Do you wak a dog? Yes No 5 Do you have someone with whom you can go for a wak, or do other physica activities? Aways Often Sometimes Never 6 Do you ride a bicyce? Yes No (pease go to question 7) 6a 6b 6c If yes, how many times did you cyce ast week?. times How ong on average did you cyce for each time?.....minutes How woud you describe your cycing pace? Sow Average Fast 7 Do you go swimming? Yes No (pease go to question 8) 7a 7b If yes, how many times did you swim ast week?..times How ong on average did you swim for each time?...minutes 138 NIHR Journas Library

171 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 7c How woud you describe your swimming speed? Sow Average Fast 8 Do you have a garden or aotment? Yes No (pease go to question 9) 8a If yes, how many hours, on average, a week do you spend doing gardening? In summer hours In winter hours 9 Have you participated in any sporting activities in the ast week? Yes No (if no, pease go to question 10) 9a If yes, what kind of sporting activity?. 9b How many hours approximatey did you spend participating in sporting activities in the ast week? Less than 1 hour in the ast week hours in the ast week 10 How often did you perspire during physica activity in the ast week? Never 1-2 times 3-4 times 5 or more times 11 Do you have a staircase in your home? Yes No 11a Do you cimb stairs reguary (at east once per day)? Yes No (go to section I) 11b If yes, approximatey how many times per day do you cimb the stairs?... times per day Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 139

172 APPENDIX 1 Section I Some questions about your beief in your abiity to exercise How sure are you that you wi do each of the foowing: 1 Exercise reguary (3 times a week for 20 minutes) 2 Exercise when you are feeing tired 3 Exercise when you are feeing under pressure to get things done 4 Exercise when you are feeing down or depressed 5 Exercise when you have too much work to do at home 6 Exercise when there are other more interesting things to do 7 Exercise when your famiy or friends do not provide any support 8 Exercise when you don t reay fee ike it 9 Exercise when you are away from home (e.g. visiting, on hoiday) Very Pretty A itte Not at a Sure Sure Sure Sure 140 NIHR Journas Library

173 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section J Finay, some questions about you & your iving circumstances 1 What is your current marita status? Married (or iving with someone as a coupe) Widowed Divorced or separated Singe Other If other, pease describe 2 How many peope in your househod, incuding yoursef, are there Aged under 18 Aged Aged 65 or over 3 Who ives in your househod with you? (pease tick a that appy) I ive on my own (pease go to question 4) My husband / wife / partner Other famiy members Other aduts 4 Do you have someone with whom you woud be abe to discuss a very persona and serious probem? Yes No 5 At what age did you finish your continuous fu-time education at schoo, coege or university? 14 or under or over Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 141

174 APPENDIX 1 More questions about you & your iving circumstances 6 Do you have any quaifications? Yes No (Pease go to question 7) If yes, which of these quaifications do you have? If you have any of the quaifications isted, pease tick every box that appies. If your UK quaifications are not isted, tick the box that contains its nearest equivaent. If you have quaifications from outside the UK, tick the Foreign quaifications box and the nearest UK equivaents (if known). 1-4 O eves / CSEs / GCSEs (any grades), Entry Leve, Foundation Dipoma NVQ Leve 1, Foundation GNVQ, Basic Skis 5+ O eves (passes) / CSEs (grade 1) / GCSEs (grades A*- C), Schoo Certificate, 1 A eve / 2-3 AS eves / VCEs, Higher Dipoma NVQ Leve 2, Intermediate GNVQ, City and Guids Craft, BTEC First / Genera Dipoma, RSA Dipoma Apprenticeship 2+ A eves / VCEs, 4+ AS eves, Higher Schoo Certificate, Advanced Dipoma NVQ Leve 3, Advanced GNVQ, City and Guids Advanced Craft, ONC, OND, BTEC Nationa, RSA Advanced Dipoma Degree (for exampe BA, BSc), Higher degree (for exampe MA, PhD,PGCE) NVQ Leve 4-5, HNC, HND, RSA Higher Dipoma, BTEC Higher Leve Professiona quaifications (for exampe teaching, nursing, accountancy) Other vocationa / work-reated quaifications Foreign quaifications No quaifications 7 What is your empoyment status? In fu time empoyment In part time empoyment Seeking work Looking after home or famiy Retired Student Not working due to ong-term sickness or disabiity Other (pease describe) 142 NIHR Journas Library

175 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Answer question 8 for your main job, or if you are not working, your ast main job. Your main job is the job in which you usuay work (worked) the most hours. 8 What is (was) your fu and specific job tite? For exampe, PRIMARY SCHOOL TEACHER, CAR MECHANIC, DISTRICT NURSE, STRUCTURAL ENGINEER Do not state your grade or pay band. 8a Briefy describe what you do (did) in your main job. 8b At your workpace, what is (was) the main activity of your empoyer or business? For exampe, PRIMARY EDUCATION, REPAIRING CARS, CONTRACT CATERING, COMPUTER SERVICING. If you are (were) a civi servant, write GOVERNMENT 9 Do you, or the peope you ive with, own or rent your own home? Own (with or without a mortgage) Rent from counci or housing association Rent privatey Other, pease describe 10 Do you have to cut back spending or borrow money to pay your eectricity, gas, teephone or counci tax bis? Aways Often Occasionay Never 11 In tota, how many cars or vans are owned, or avaiabe for use, by members of your househod? None One Two Three Four or more 11 Do you yoursef drive a car or van? Never Occasionay Most days Every day Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 143

176 APPENDIX 1 12 What is your ethnic group? Pease choose one section from A to E, then tick group or background. A White Engish / Wesh / Scottish/ Northern Irish / British Irish Gypsy or Irish Traveer Any other White background,write in.. one box to best describe your ethnic B Mixed /mutipe ethnic groups White and Back Caribbean White and Back African White and Asian Any other Mixed /mutipe ethnic background, write in. C Asian / Asian British D Back / African / Caribbean / Indian Back British Pakistani African Bangadeshi Caribbean Chinese Any other Back / African / Caribbean Any other Asian background,write in background, write in.... E Other ethnic group Arab Any other ethnic group, write in Pease write beow any other comments you have on your heath or this questionnaire Thank you for fiing in this questionnaire. 144 NIHR Journas Library

177 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 study Heath and ifestye 3 month survey Study IDNO Thank you for fiing in this questionnaire. It wi take you about 15 minutes to compete. Pease fee free to write comments by any question. A information wi be kept stricty confidentia. Pease enter your date of birth Pease enter today s date / / / / Thank you Usua activity group Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 145

178 APPENDIX 1 Section A - Some genera questions about your heath Pease put a tick in the box next to the most appropriate answer for each question. 1 How is your heath in genera? Very good Good Fair Poor Very poor 2 How much physica or bodiy pain have you had in the past 4 weeks? None Very mid or mid Moderate Severe or very severe 3 In the past four weeks, how much did pain interfere with your norma activities? Not at a A itte bit Moderatey Quite a bit or extremey 146 NIHR Journas Library

179 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section B - Questions about your heath today Under each heading, pease tick the ONE box that best describes your heath TODAY 1 Mobiity I have no probems in waking about I have sight probems in waking about I have moderate probems in waking about I have severe probems in waking about I am unabe to wak about 2 Sef-care I have no probems with sef-care I have sight probems washing or dressing mysef I have moderate probems washing or dressing mysef I have severe probems washing or dressing mysef I am unabe to wash or dress mysef 3 Usua activities (e.g. work, study, housework, famiy or eisure) I have no probems doing my usua activities I have sight probems doing my usua activities I have moderate probems doing my usua activities I have severe probems doing my usua activities I am unabe to do my usua activities 4 Pain / discomfort I have no pain or discomfort I have sight pain or discomfort I have moderate pain or discomfort I have severe pain or discomfort I have extreme pain or discomfort 5 Anxiety / depression I am not anxious or depressed I am sighty anxious or depressed I am moderatey anxious or depressed I am severey anxious or depressed I am extremey anxious or depressed Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 147

180 APPENDIX 1 Section B - Continued: about your heath today We woud ike to know how good or bad your heath is TODAY The scae is numbered 0 to means the best heath you can imagine 0 means the worst heath you can imagine Mark an X on the scae to indicate how your heath is TODAY Now, pease write the number you marked on the scae in the box beow YOUR HEALTH TODAY = Section C - Some questions on injuries and heath These questions ask about any injuries or changes in your heath that you may have had in the 3 months that you have been invoved in this study. In the ast 3 months have you had any of the foowing: 1 A fa? Yes No 1a If yes, how many times?....times in the ast 3 months 2 Any fractures (broken bones)? Yes No 2a If yes, pease give detais of what bones were injured 3 Any sprains or injuries? Yes No 3a If yes, pease give detais of the sprain or injury 148 NIHR Journas Library

181 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Some questions on injuries and heath continued If you have not had a fa, fracture, sprain or injury, pease go to question 6. If you have had a fa, fracture, sprain or injury, pease go to question Did you or your famiy have to pay for anything as a resut of your fa(s), fracture(s), sprain(s) or injury (ies)? (Pease consider any costs inked to your continuing care or recovery) Yes No 4a. If yes, roughy how much did you spend?... 4b What was this spent on In the past 3 months did you have to stop doing your usua activities due to a fa, fracture, sprain or injury? Yes No 5a. If yes, how many days did you stop your usua activities?..days 6 In the ast 3 months have you noticed a deterioration in any heath probems that you aready had at the start of this research project? Yes No 6a If yes, pease give detais 7 In the ast 3 months have you noticed an improvement in any heath probems that you aready had at the start of this research project? Yes No 7a If yes, pease give detais Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 149

182 APPENDIX 1 Section D - Some questions on how you fee For each item beow, pease tick the box opposite the repy that comes cosest to how you have been feeing over the past week. Don t take too ong over the answers: your immediate reaction wi probaby be most accurate. Tick ony one box in each section 16. I fee tense or wound up : Most of the time A ot of the time Time to time Not at a 17. I fee as if I am sowed down: Neary a of the time Very often Sometimes Not at a 18. I sti enjoy things I used to: Definitey as much Not quite as much Ony a itte Hardy at a 19. I get a sort of frightened feeing ike butterfies in the stomach: Not at a Occasionay Quite often Very often 20. I get a sort of frightened feeing as if something bad is about to happen: Very definitey Yes, but not too bady A itte, but it doesn t worry me Not at a 21. I have ost interest in my appearance: Definitey I don t take so much care as I shoud do I might not take quite as much care I take just as much care 22. I can augh and see the funny side of things: As much as I aways coud Not quite so much now Definitey not so much now Not at a 150 NIHR Journas Library

183 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO I fee restess, as if I have to be on the move Very much indeed Quite a ot Not very much Not at a 24. Worrying thoughts go through my mind: A great dea of the time A ot of the time From time to time but not too often Ony occasionay 25. I ook forward with enjoyment to things: As much as I ever did Rather ess than I used to Definitey ess than I used to Hardy at a 26. I fee cheerfu: Not at a Not often Sometimes Most of the time 27. I get sudden feeings of panic Very often indeed Quite often Not very often Not at a 28. I can sit at ease and fee reaxed: Definitey Usuay Not often Not at a 29. I can enjoy a good book, radio or TV programme: Often Sometimes Not often Very sedom 30. I fee oney: A the time Often Sometimes Never Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 151

184 APPENDIX 1 Section E - Some questions about your attitudes to exercise and heath Pease indicate how strongy you agree or disagree with each statement Pease tick one box to indicate how strongy you agree or disagree with each statement Strongy agree Sighty agree Unsure Sighty disagree Strongy disagree 7. Doing exercise is satisfying and rewarding to me 8. Doing exercise reguary is good for me 9. There is itte I can do to make up for the physica osses that come with age 10. Exercising reguary can be hepfu for my heath 11. Exercising reguary can hep me to get out of doors 12. Exercising reguary can hep me to contro my weight or to ose weight 152 NIHR Journas Library

185 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 Section F Some questions about your beief in your abiity to exercise How sure are you that you wi do each of the foowing: 1 Exercise reguary (3 times a week for 20 minutes) 2 Exercise when you are feeing tired 3 Exercise when you are feeing under pressure to get things done 4 Exercise when you are feeing down or depressed 5 Exercise when you have too much work to do at home 6 Exercise when there are other more interesting things to do 7 Exercise when your famiy or friends do not provide any support 8 Exercise when you don t reay fee ike it 9 Exercise when you are away from home (e.g. visiting, on hoiday) Very Pretty A itte Not at a Sure Sure Sure Sure Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 153

186 APPENDIX 1 Section G - Some questions about physica activity 6 How many times did you take a wak outside during the ast week? (incude waking reated to other activities)...times ast week 2 How ong did such a wak usuay ast?..minutes 3 Did you take a wak that asted onger than 1 hour during the ast month? Yes No 3a If yes, how many times did you do that?.times ast month 4 Do you have someone with whom you can go for a wak, or do other physica activities? Aways Often Sometimes Never 5 Do you ride a bicyce? Yes No (pease go to question 6) 5a 5b 5c If yes, how many times did you cyce ast week?. times How ong on average did you cyce for each time?.....minutes How woud you describe your cycing pace? Sow Average Fast 6 Do you go swimming? Yes No (pease go to question 7) 6a If yes, how many times did you swim ast week?..times 6b How ong on average did you swim for each time?...minutes 6c How woud you describe your swimming speed? Sow Average Fast 154 NIHR Journas Library

187 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO Have you participated in any sporting activities in the ast week? Yes No (if no, pease go to question 8) 7a If yes, what kind of sporting activity?. 7b How many hours approximatey, did you spend participating in sporting activities in the ast week? Less than 1 hour in the ast week hours in the ast week 8 How often did you perspire during physica activity in the ast week? Never 1-2 times 3-4 times 5 or more times 9. In the ast 3 months (since you have been taking part in the PACE-UP tria) do you think that your waking and physica activity has: Decreased a ot Decreased a itte Stayed about the same Increased a itte Increased a ot Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 155

188 APPENDIX 1 Section H - Some questions about the money you have spent to do with waking and other physica activity 7 In the past 3 months, did you pay for any membership fees to do with waking? No (pease go to question 2) Yes (pease go to question 1a) 1a. If yes, how much did you spend?.. 1b. How often do you tend to pay this amount? (circe the correct frequency beow) Weeky / Monthy/ Annuay / Each time / It is a one off / Don t know / Other If other, pease specify.. 2. In the past 3 months, did you pay for any individua casses, entrance fees or groups to do with waking? (if not incuded in membership fees above) No (pease go to question 3) Yes (pease go to question 2a) 2a. If yes, how much did you spend?.. 2b. How often do you tend to pay this amount? (pease circe the correct frequency beow) Weeky / Monthy/ Annuay / Each time / It is a one off / Don t know / Other If other, pease specify.. 3. In the past 3 months, did you pay for shoes or cothing to do with wak No (pease go to question 4) Yes (pease go to question 3a) 3a. If yes, how much did you spend?.. 3b. How often do you tend to pay this amount? (pease circe the correct frequency beow) Weeky / Monthy/ Annuay / Each time / It is a one off / Don t know / Other If other, pease specify. 156 NIHR Journas Library

189 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO In the past 3 months, did you have to pay for food or drink to do with waking? No (pease go to question 5) Yes (pease go to question 4a) 4a. If yes, how much did you spend?.. 4b. How often do you tend to pay this amount? (pease circe the correct frequency beow) Weeky / Monthy/ Annuay / Each time / It is a one off / Don t know / Other If other, pease specify.. 5. In the past 3 months, did you have to pay for anything ese to do with waking? No (pease go to question 6) Yes (pease go to question 5a) 5a. If yes, what ese did you have to pay for?.. 5b. How much did you spend?.. 5b. How often do you tend to pay this amount? (pease circe the correct frequency beow) Weeky / Monthy/ Annuay / Each time / It is a one off / Don t know / Other If other, pease specify.. 6. In the past 3 months, did you spend money on other kinds of physica activity? No Yes (pease go to question 5a) 6a. If yes, what other kinds of physica activity did you spend money on? (pease ist a that appy)... 6b. If yes, roughy how much did you spend in tota on other kinds of physica activity over the past 3 months? In the past 3 months do you think that your spending on waking and physica activity has: Increased a ot Increased a itte Stayed about the same Decreased a itte Decreased a ot Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 157

190 APPENDIX 1 Pease write beow any other comments you have on your heath or this questionnaire Thank you for fiing in this questionnaire. 158 NIHR Journas Library

191 DOI: /hta22370 HEALTH TECHNOLOGY ASSESSMENT 2018 VOL. 22 NO. 37 study Heath and ifestye 3 month survey Study IDNO Thank you for fiing in this questionnaire. It wi take you about 15 minutes to compete. Pease fee free to write comments by any question. A information wi be kept stricty confidentia. Pease enter your date of birth Pease enter today s date / / / / Thank you Pedometer by post group Queen s Printer and Controer of HMSO This work was produced by Harris et a. under the terms of a commissioning contract issued by the Secretary of State for Heath and Socia Care. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 159